Global warming has intensified the frequency and severity of extreme weather events, which degrade CCTV signal and video quality while disrupting traffic flow, thereby increasing traffic accident rates. Existing datasets, often limited to light haze, rain, and snow, fail to capture extreme weather conditions. To address this gap, this study introduces the Traffic Surveillance Benchmark for Occluded vehicles under various Weather conditions (TSBOW), a comprehensive dataset designed to enhance occluded vehicle detection across diverse annual weather scenarios. Comprising over 32 hours of real-world traffic data from densely populated urban areas, TSBOW includes more than 48,000 manually annotated and 3.2 million semi-labeled frames; bounding boxes spanning eight traffic participant classes from large vehicles to micromobility devices and pedestrians. We establish an object detection benchmark for TSBOW, highlighting challenges posed by occlusions and adverse weather. With its varied road types, scales, and viewpoints, TSBOW serves as a critical resource for advancing Intelligent Transportation Systems. Our findings underscore the potential of CCTV-based traffic monitoring, pave the way for new research and applications. The TSBOW dataset is publicly available at the following link.

Ultra Thin Glass (UTG) is a key component of foldable electronics due to its excellent flexibility and strength. However, water droplet remaining between UTG and vacuum stage after strength process disturbs vision systems. Such systems often perceive droplets as cracks, resulting in lower production efficiency. Due to the limitations of the system and budget, solutions with deep learning are challenging. The work presents a non-deep learning solution to adjust the surface roughness of vacuum stages. This improves defect classification between water droplet and real defect. A test environment was built using a 6axis robot, an air knife, and two ccd vision sets. Two stages (Ra ≈ 3.5 and Ra ≈ 5.0) were compared and tested. The rough stages had a higher water removal rate, and the over-detection rate was reduced to less than 1%. Studies show that even basic surface adjustments can increase inspection precision without complex AI. This provides an affordable and practical solution in the current production environment.

Traffic flow estimation is an essential task in modern intelligent transportation systems. Many types of information, including vehicle type, vehicle totals, and movement direction, are vital for mitigating transportation-related tasks and effective traffic control strategies. With the development of embedded devices, systems can process captured video at the edge instead of transferring data to centralized processing servers. This paper proposes a real-time and edge-based traffic flow estimation system. The proposed system follows a detect-and-track mechanism where lightweight deep learning models perform vehicle detection. A novel scenario-based tracking and counting technique is developed to provide multi-class, multi-movement vehicle counting. The method uses predefined regions to assign the movement for each vehicle initially detected. It then performs spatial-temporal trajectory matching between the vehicle trajectory and the movement path throughout the whole video. Extensive experiments have shown that the proposed method achieves high effectiveness with multiple camera types and viewpoints.

The study in this paper aims to demonstrate the performance of a junction detection architecture based on a deep learning paradigm, included with dark channel transformation. We must take into account the hostile conditions when detecting such features. Many of previous papers proposed models for junction detection with hand-crafted logic, which works well under normal conditions but not under hostile conditions. This necessitates a data-driven approach for junction detection. We attempt to do so using two recently proposed deep neural networks: ResNet50 and EfficientNet-B0. Here, given a set of input images of roads with junctions or no junctions, dark channel transformation is applied to better inform the networks about the road regions prominent in the images. According to our experiments on the Oxford RobotCar Dataset, using the dark channel transformation on the ResNet50 trained from scratch can achieve a junction classification accuracy of over 94%. This numerical figure is 8% greater than when the ResNet50 pre-trained on ImageNet is directly trained on RGB Oxford dataset images. When using pre-trained weights of the deep networks, junction classification accuracy rises to 95%, and the precision increases to 99%.

In this paper, we propose a method to perform training and inference with multiple instances of the same deep neural network architecture on images taken from cameras of different directions. Across multiple cameras, depending on each of their directional characteristics, objects viewed from the cameras can form slightly different distributions in visual features. Regarding this, we emphasize the importance of camera-wise training on multiple instances of a given deep neural network for object detection. Given the Waymo Open Perception Dataset, we used multiple instances of the YOLOv5x6 architecture and trained each of them per camera. Such a training scheme on the Training Set achieves better training progression, and the inference results are shown to have AP/L1 as high as 0.6679 on the Testing Set.

In this paper, we present a deep learning model for high-accuracy, high-speed detection of vehicle taillights in traffic. The model consists of three major modules: the lane detector, the car detector, and the taillight detector. Unlike most previously proposed algorithms where hand-coded schemes are used, we have adopted a data-driven approach. This data-driven scheme was implemented in both the car and taillight detection modules. First, we used an intricately designed lane detection module, then we adopted the Recurrent Rolling Convolution (RRC) architecture and tracking mechanism for detecting car boundaries. Subsequently, we used the same RRC architecture to extract the taillight regions of the detected cars. The lane detection and car detection modules improve both the speed and detection rate of the final taillight detection. The robustness of our model was verified using datasets from Sungkyunkwan University (SKKU) as well as the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI). Our model works well even in hostile conditions. It achieves detection rates as high as 99% in testing with the SKKU dataset. When using the KITTI 2D Object dataset, the model achieves a taillight detection rate of 86%. The model achieves 100% taillight detection rate on a certain, small subset of the KITTI Tracking dataset.

While it is a hot issue whether cars could drive by themselves in emergent situations without any kind of human interference, pedestrian detection is the key technology in autonomous driving cars. Though current pedestrian detection technologies have come to a point in which they are accurate in normal conditions and surroundings, existent systems are inaccurate in harsh situations, such as when there are too many pedestrians, when there is too much light or when it is too dark, or when it is raining or snowing heavily. This problem may be solved by integrating deep learning and combining a new type of local pattern with the RGB raw image as input, instead of using just the RGB image as input. We will introduce a new type of local pattern called Triangular Patterns, which is effective in extracting more detailed and stable features from local regions. Here in this paper, we propose a pedestrian detection system in which deep learning is used, along with combining the RGB raw image with Triangular Patterns for input.

The objective of this study was to propose universal edge gateway that can transfer information of various legacy equipment conveniently and cost effectively to IT system for smart factory. The proposed universal edge gateway converts information from various legacy equipment used at the manufacturing site to OPC UA protocol. Also, ISA 95 information model was applied to realize the same equipment interface for all IT systems, including MES and ERP. It also includes hardware that can connect additional sensors, required for legacy equipment and configuration tools that allow operators to use simple gateways without modifying the equipment. As a result, in IT systems, it is possible to access the information of legacy equipment easily with the same interface, and to reduce the cost and time for building smart factory.

Change/motion detection is a challenging problem in video analysis and surveillance system. Recently, the state-of-the-art methods using the sample-based background model have demonstrated astonishing results with this problem. However, they are ineffective in the dynamic scenes that contain complex motion patterns. In this paper, we introduce a novel data-driven approach that combines the sample-based background model with a feature extractor obtained by training a triplet network. We construct the network by three identical convolutional neural networks, each of which is called a motion feature network. Our network can automatically learn motion patterns from small image patches and transform input images of any size into feature embeddings for high-level representations. The sample-based background model of each pixel is then employed by using the color information and the extracted feature embeddings. We also propose an approach to generate triplet examples from CDNet 2014 for training our network model from scratch. The offline trained network can be used on the fly without re-training on any video sequence before each execution. Therefore, it is feasible for real-time surveillance systems. In this paper, we show that our method outperforms the other state-of-the-art methods on CDNet 2014 and other benchmarks (BMC and Wallflower).

The rectification process is a compulsory step in stereo matching computation. To obtain depth information, stereo camera systems are often installed in vehicles for outdoor and street-related applications, including vehicle and pedestrian detection, lane detection, and traffic sign recognition. In this paper, we propose a rectification method that uses currently available front- and rear-view vehicle cameras to produce rectified stereo images. The proposed method can be employed with different types of cameras that have varying focal lengths. In addition, this method tolerates the problem of camera alignment variation from normal stereo camera systems. To achieve this, a compensation method for different focal lengths and the estimation of image relationships are introduced. The experimental results demonstrate that the proposed method can operate robustly and accurately with different kinds of stereo images and significantly outperforms a state-of-the-art rectification method.

Traditional deep learning-based vehicle detection methods are often designed using a pyramid of filters with multiple scales and sizes; therefore, the processing time is slow due to the large number of scales used and because the classifier runs at all scales. Recently, a deep learning-based region proposal network was introduced to detect vehicles that only employ the network one time regardless of the size of the input image. In object detection, deep learning-based region proposal networks have achieved state-of-the-art performance in terms of accuracy. These systems achieve a very high accuracy under normal driving conditions; however, their performance decreases under difficult driving conditions such as in snow, rain, or fog. In addition, the current state-of-the-art system-based region proposal networks still fail to satisfy the real-time requirements of the driving assistant systems. More recently, the identification of local patterns has been shown to improve the performance of the traditional deep-learning systems; hence, this paper investigates local patterns in region proposal networks to improve their accuracy. Depth information is also investigated to improve the processing time of current region proposal networks. Our experimental results show that the proposed system obtains better performance than the state-of-the-art object region detection systems in terms of both accuracy and processing time.

Car detection is a crucial issue in self-driving cars. Numerous in-traffic car detection models have been proposed, each of which exhibits its own strengths and weaknesses; the high detection speeds of some models are not accompanied by high precision, while the precision of other models is shadowed by insufficient speeds. Our main goal in this paper is to introduce a model that utilizes the Recurrent Rolling Convolution (RRC). The model gives promising results on detection speed and precision, thereby mitigating the weaknesses of previously proposed models, which is exhibited in our extensive experiment.

In this paper, we propose a method to improve the performance of the fiducial mark detection function using a vision sensor in automation equipment. In the automation industry, Template matching method is used to recognize the fiducial mark. Template matching can be detected because the error increases when the mark of the target rotates more than a certain angle. If the mark is damaged due to illumination and a physical external force, there is a reduction in the recognition rate. Therefore, we propose a method consisting of K-means Clustering, SVM Classification, and Linear Regression. Using the proposed method, the recognition rate of the fiducial mark is improved and accurate center coordinates are obtained.

With the emergence of industry 4.0, autonomous driving vehicles have become an exciting research topic in the science technology community. The driving system requires many complex algorithms that provide both accurate results and fast running times. However, the performance of these algorithms is usually limited to standard personal computer (PC) systems. Currently, the computation power of available embedded systems lags far behind that of a standard PC, even when compared to PCs with moderate capabilities. Hence, we present some benchmark results of several systems, including standard PCs, laptops, and embedded systems, for performing computer vision algorithms. We collected a set of algorithms that are commonly used in autonomous driving systems and then ran each of these on our selected systems. In this evaluation, we focused only on the processing time without concerning about the precision of the algorithm. The details of the testing algorithms and systems are provided in this report also. We believe that our experiment can provide practical information to people who aim to transfer their algorithm to an embedded system.

This paper addresses issues that occurs when using an asynchronous serial communication mode instead of encoder communication, such as (bidirectional/serial/synchronous) BiSS communication in an encoder system, and operation without using a slave module when using BiSS communication. Therefore, this paper discusses how to operate using a microcontroller unit (MCU), a field-programmable gate array (FPGA), and a RS422 transceiver chip without using a slave module. It also explains the problems that arise when using the MCU and solves the problem by using the FPGA. The effectiveness of the proposed method is demonstrated by the experimental results.

Pedestrian detection is fundamental challenge for computer vision which requires localizing objects within an image. Convolutional neural networks are widely used in object recognition. However, ordinal convolutional methods using sliding window as the input for networks require time to run an entire image and can only handle a fixed size window image. We propose to using a region proposal based in the V-disparity method to obtain prospect regions, instead of the original scanning methods to obtain the object regions. The region proposals from the V-disparity will be fed as the input for a convolutional neural network(CNN). We also extend the CNN for more effective task object detection. In our model, CNN are combined with recursive neural networks to learn features and classify color images. The convolutional neural network layer learns low-level features of the input image. By using CNN, the learned features can represent highly variable objects in the input image. The learned features after the convolutional layer are then given as inputs to a recursive neural network (RNN) to compose higher order features. The RNN is a multiple, fixed-tree recursive which can combine convolution and pooling into one efficient hierarchical operation.

This paper introduces a manipulator-position control procedure for which the Kinect sensor is used with 3D-modeling data. The system consists of the Manipulator Control software and the Kinect sensing software. The Manipulator Control software is used to control the AT2, an industrial 6-DOF manipulator. The Kinect sensing software is used to obtain the positional data of the object whose 3D-model data is registered in advance. The manipulator-control software obtains the position data from the Kinect sensing software using a Windows message. A median filter is used to remove the errors from the positional data. The rotational matrix is calculated for the coordinate transformation. The difference between the scales and translations of the axes are compensated by using the coordinates of both systems at two different positions. Real-system experiment results are presented to prove the validity of the proposed method.

The signals of magnetic encoders (MEs) are a pair sinusoidal waves: a sine wave and a cosine wave. Due to many different causes, they contain error factors such as offsets, amplitude differences, phase shifts, and random noise. This paper proposes auto-calibration and noise-reduction methods to reduce these errors and to estimate the positions of MEs. This method includes auto-calibration and adaptive-bandwidth phase-locked loop (ABW-PLL) algorithms. The auto-calibration algorithm normalized the sinusoidal signals of the MEs to the range of [-1; 1]. The ABW-PLL algorithm accurately estimated the phase angle of the MEs. Unlike the traditional filters, the ABW-PLL almost eliminated the time lag self-adaptively based on the input signals. The effectiveness of the proposed method is demonstrated by the simulation and experiment results.

This paper presents the design and implementation of a motion control system using EtherCAT. The EtherCAT frame with an embedded motion profile is introduced. Additionally, a closed loop stepper motor that uses the embedded data of the EtherCAT frame is shown. For position control purposes, a lead angle estimation was designed. The Proportional Integral Feed Forward (PI-FF) controller was implemented to overcome the problem with reverse motor rotation. Finally, some experimental results are shown to verify the performance of the EtherCAT based motor drive.

This paper presents an approach for the improvement of the accuracy of absolute magnetic encoders (AME). The encoders comprise the following two magnets: a multipolar magnet to increase the resolution and the accuracy, and a center-located bipolar magnet for the calculation of the absolute angle. The multipolar signal processing is crucial for the increasing of the encoder accuracy; however, the multipolar signals are not ideal, i.e, dc offsets, different amplitudes, phase shifts, and random noise. The present paper proposes a calibration method that is based on the adaptive linear neural network (ADALINE) for the reduction of the effect of the nonidealities. In addition, to optimize the loop-acquisition time and to enhance the random-noise reduction, the bandwidth is adapted using the fuzzy phase-locked-loop (F-PLL). This method is simulated using Matlab software and is implemented on the ARM STM32F405R. The study results demonstrate the efficient high performance that can be achieved with the use of the proposed method.

In Network Control Systems (NCSs), a dual rate controller is an approach that is applied to reduce the influence of packages loss and random delays. If the output measuring sensors is slower than the control action update can be reached, the dual rate controller is used as a possible solution. In this paper, a fuzzy controller will be combined with a dual rate controller in an NCS to regulate the gain parameter online. The fuzzy dual rate PID is designed by slitting a convention PID into two parts performing at different sampling rate. A Controller Area Network (CAN) is the mode of communication in this system. Furthermore, the Fuzzy-dual rate PID controllers performance is compared with that of a conventional PID controller. Based on the simulation results and experimental results, the fuzzy dual rate PID controller provides a superior position response with a faster rise time, faster settling time and it minimizes the amount of overshoot.

Recently, Real-time Ethernet has become popular communication technology in automation system applications due to its high bandwidth capacity and low market cost. In addition, CC-Link IE Field has been successful in combining the ultra-high speed feature with the large amount of data to be transferred while integrating in field-level industrial communication networks. This protocol also provides a seamless structure that provides the ability to implement a gateway to communicate with different protocols. On the other hand, many devices such as motor driver still apply CAN or serial RS-485 in automation systems, which means that CC-Link IE Field requires time and cost to allow these devices to implement this new protocol. In order to alleviate this problem, this paper proposes a method of designing a gateway for communicating between the CC-Link IE Field and serial RS-485 based protocol to increase the adoption rate of using CC-Link IE Field in automation system networks. This gateway structure was implemented and verified in experiments of using CC-Link IE Field Master to control motor drives, thus supporting the RS-485 communication interface.

A real-time hardware architecture based on scale-invariant feature transform algorithm (SIFT) feature extraction with parallel technology has been introduced in this paper. The proposed parallel hardware architecture could be able to extract feature via a Field-Programmable Gate Array (FPGA) chip efficiently, which provided the real-time performance and the similar accuracy with software implementation. In terms of hardware resource consumption and speed, the original SIFT algorithm has been significantly optimized in the following aspects: 1) Down-sampling has used to replace with up-sampling for purpose of saving the interpolation calculation. Besides, the flexible Gaussian blur value and self-circulation scale space are proposed to simplify the key point detection operation. 2) Optimized key point detection method replaces the original key point detection methods for there won't be other key points in the neighbor 8 pixels in the same scale if a certain pixel were defined as a key point. Compared with the SIFT algorithm implemented in the software, this kind of architecture based on FPGA performance is similar with software methods and realizes the efficient and real-time.

Stereo matching is a challenging task because stereo images are affected by many factors, such as radiometric distortion, sun and rain flare, flying snow, occlusion, textureless and noisy image regions, and object boundaries. However, most of the existing methods for stereo matching aim to solve only one specific problem. As a result, their performance is degraded significantly when operating with stereo images captured under a variety of scenes and conditions. In this paper, we propose a novel matching cost function based on adaptive normalized cross-correlation (ANCC). We demonstrate several weaknesses of ANCC and propose techniques to resolve them. In addition, we employ available information, such as intensity mean, intensity variance, and support window radius, to estimate the parameters of the proposed matching cost function. Compared with ANCC, the proposed matching cost function reduces the error rates from 24.1% to 17.8% in the Middlebury data set and from 64.1% to 26.4% in the KITTI data set. In addition, for noisy stereo pairs, the proposed function reduces the error rate from 73.6% to 37.3%. The qualitative and quantitative experimental results based on stereo images in different data sets under various conditions show that our proposed matching cost function outperforms state-of-the-art matching cost functions in indoor and outdoor stereo images having various radiometric distortions.

The performance of stereo matching algorithms strongly depends on the quality of the stereo data/matching cost. Most state-of-the-art data costs require expert knowledge for the design of a transformation function, such as census for handling gray-level changes monotonically, adaptive normalized cross correlation for handling Lambertian cases, guided filtering for preserving edge information, and local density encoding for handling illumination differences. However, it is difficult to design a complex transformation function to handle unknown factors that often occur in driving conditions such as snow, rain, and sun. Therefore, this paper has investigated the deep learning strategy to develop a novel stereo matching cost model without using much expert knowledge. Experimental results show that the proposed deep learning model obtains better results than the state-of-the-art stereo matching cost as judged by the standard KITTI benchmark, Middlebury, and HCI datasets.

This paper introduces a general framework for detection, recognition, and tracking preceding vehicles and pedestrians based on a deep learning approach. The proposed framework combines a novel deep learning approach with the use of multiple sources of local patterns and depth information to yield robust on-road vehicle and pedestrian detection, recognition, and tracking. The proposed system is first based on robust obstacle detection to identify obstacles appearing along the road that are likely to be vehicles and pedestrians, implemented as an efficient adaptive U-V disparity algorithm. Second, the results from the obstacle detection stage are input into a novel vehicle and pedestrian recognition system based on a deep learning model that processes multiple sources of depth information and local patterns. Finally, the results from the recognition stage are used to track detected vehicles or pedestrians in the next frame by means of a proposed tracking and validation model. The proposed framework has been thoroughly evaluated by inputting several vehicle and pedestrian data sets that were collected under various driving conditions. Experimental results show that this framework provides robust vehicle and pedestrian detection, recognition, and tracking with high accuracy, and also satisfies the real-time requirements of driver assistance systems.

This work presents a method for automating the glass cutting process in a product line of a touch screen panel. The mechanical roller cutting method has been used for some time for industrial processes of glass. However, the existing mechanical glass cutting method needs workers to break the glass after the scribing process. The glass can become damaged during the breaking process because it is difficult for a human to apply the same pressure and direction every time when breaking the glass. If we automate these processes, we can apply consistent pressure and breaking angle. As a result, we can obtain an improved production yield rate. In addition, the breaking process can be dangerous for worker operating the machinery. Glass micro powders scatter during the scribing and breaking process. When workers handle the glass, they are exposed to this danger. Therefore, an automatic glass cutting machine needs to be developed in order to eliminate workers from these processes.

Stereo correspondence is a challenging task because stereo images are affected by many factors such as radiometric distortion, sun and rain flares, flying snow, occlusions and object boundaries. However, most of the existing stereo correspondence methods use simple matching cost functions. As a result, their performance is degraded significantly when operating with real‐world stereo images whose intensities of corresponding pixels can be arbitrarily transformed. In this study, the authors propose a novel matching cost function based on the order relations between pixel pairs that can operate accurately under various conditions of transformed intensities between stereo images. The proposed matching cost function is an improvement of the soft rank transform (SRT) and can tolerate local, monotonically non‐linear changes in intensities between the left and right images. The proposed function significantly reduces the error rate from 24.7 to 12.7% in the Middlebury dataset, and from 19.8 to 7.1% in the KITTI dataset as compared with the SRT. The qualitative and quantitative experimental results obtained using stereo images in different datasets under various conditions show that the proposed matching cost function outperforms the state‐of‐the‐art matching cost functions in indoor and outdoor stereo images.

We propose a novel fuzzy encoding pattern that fuzzily encodes the relative orders between pixel pairs. An image window is divided into disjoint neighboring pixel sets for the window's center pixel, and the relative order is established not only between the center pixel and its neighbors but also between the pixel pairs in each neighboring pixel set. The relative orders are fuzzily encoded to extract more detailed information from a local structure. We successfully apply the pattern as a matching cost function for stereo correspondence under severe radiometric variations. We conduct experiments using the proposed matching cost function and compare it with functions employing the census transform, supporting local binary pattern, and adaptive normalized cross correlation, as well as a mutual information-based matching cost function, using different stereo data sets. Compared with the census transform, the proposed function reduces the error from 33.1% to 16.9% in the Middlebury data set and from 17.6% to 9.5% in the Kitti data set. The experimental results indicate that the proposed function is superior to the state-of-the-art functions under radiometric variations. In addition, the proposed function is faster than recently developed functions, such as the adaptive normalized cross correlation, a mutual information-based function, and support local binary pattern.

In this study, an advanced variational model is presented for problem modelling in computer vision and image processing. The proposed model allows for the definition of multiple constraints in data fidelity, which has not been considered in previous state‐of‐the‐art methods. With this definition, the model is more robust and flexible with regard to problem modelling. Two algorithms are introduced to solve the optimisation problems: one for the vector domain and the other for the frequency domain. The issue of multiple L 1 ‐norms in the data fidelity term is resolved with these algorithms; this remained unsolved in previous research because of the difficulty with optimisation. The proposed model is demonstrated through two problems in image processing: image denoising and image deblurring. The results indicate that, compared to previous methods, images of high visual quality were both produced and recovered when using the proposed model. In addition, good and stable results in real‐world images were yielded by the proposed model, which indicates vast potential for practical uses.

Enhancing the sharpness and reducing the noise of blurred, noisy images are crucial functions of image processing. Widely used unsharp masking filter‐based approaches suffer from halo‐artefacts and/or noise amplification, while noise‐ and halo‐free adaptive bilateral filtering (ABF) is computationally intractable. In this study, the authors present an efficient sharpening algorithm inspired by guided image filtering (GF). The author's proposed adaptive GF (AGF) integrates the shift‐variant technique, a part of ABF, into a guided filter to render crisp and sharpened outputs. Experiments showed the superiority of their proposed algorithm to existing algorithms. The proposed AGF sharply enhances edges and textures without causing halo‐artefacts or noise amplification, and it is efficiently implemented using a fast linear‐time algorithm.

Real-world stereo images are inevitably affected by radiometric differences, including variations in exposure, vignetting, lighting, and noise. Stereo images with severe radiometric distortion can have large radiometric differences and include locally nonlinear changes. In this paper, we first introduce an adaptive orthogonal integral image, which is an improved version of an orthogonal integral image. After that, based on matching by tone mapping and the adaptive orthogonal integral image, we propose a robust and accurate matching cost function that can tolerate locally nonlinear intensity distortion. By using the adaptive orthogonal integral image, the proposed matching cost function can adaptively construct different support regions of arbitrary shapes and sizes for different pixels in the reference image, so it can operate robustly within object boundaries. Furthermore, we develop techniques to automatically estimate the values of the parameters of our proposed function. We conduct experiments using the proposed matching cost function and compare it with functions employing the census transform, supporting local binary pattern, and adaptive normalized cross correlation, as well as a mutual information-based matching cost function using different stereo data sets. By using the adaptive orthogonal integral image, the proposed matching cost function reduces the error from 21.51% to 15.73% in the Middlebury data set, and from 15.9% to 10.85% in the Kitti data set, as compared with using the orthogonal integral image. The experimental results indicate that the proposed matching cost function is superior to the state-of-the-art matching cost functions under radiometric variation.

Stereo correspondence is challenging under realistic conditions due to uncontrolled factors that affect input images, including illumination inconsistencies and radiometric variations. Many local and global models have been suggested to address these problems; however, their performance is often degraded due to the assumption of color consistency between the left and right images. Therefore, we present a new local pattern, local density encoding, for stereo matching measurements to improve the performance of existing stereo methods. Our experimental results indicate that the proposed method is less sensitive to illumination changes and radiometric variations. Moreover, in the cases with normal and severe illumination changes, the proposed method is more robust than state-of-the-art data costs.

The local binary pattern (LBP) and its variants have been widely investigated in many image processing and computer vision applications due to their robust ability to capture local image structures and their computational simplicity. The existing LBPs extract local structure information by establishing a relationship between the central pixel and its adjacent pixels. However, most LBPs miss the relationship among all of the pixels in the local region. Therefore, this paper proposes a novel model to establish this relationship by introducing a support LBP. The proposed model improves the performance of the existing LBP methods and results in lower sensitivity to illumination changes and radiometric variations. Moreover, the proposed model has been successfully investigated in two applications: disparity map generation and texture classification. For disparity map generation, the proposed model reduces the root mean square (RMS) error by 23.6% (in Baby1 dataset, Middlebury), and 16.58% (in Aloe dataset, Middlebury) as compared with the standard LBP under radiometric variation conditions. Moreover, the proposed model reduces the RMS by 28.11% as compared with the standard LBP under the Gaussian noise condition in the ESATS dataset. For texture classification applications, the proposed model improves the classification results from 96.26% to 98.13% on the Outext database, from 88.03% to 91.41% on the Xu database, and from 94.00% to 96.67% on the KTH-TIPS database as compared with the completed LBP.

Mechatrolink-III is a real-time Ethernet protocol designed to achieve high performance with short cycle time. The Mechatrolink-III protocol includes five communication phases: synchronization, cyclic communication, retry communication, C1 master message communication, and C2 master message communication. These communication phases directly affect the cycle time of Mechatrolink-III. This paper analyzes the real-time performance of the Mechatrolink-III protocol according to cycle time on line and tree topologies. Moreover, we measure parameters that influence these communication phases in our system, and construct a formula to calculate the cycle time based on those parameters.

This paper presents a method to improve the accuracy of synchronous control systems for dual linear stages driven by linear servomotors in order to quickly eliminate the velocity and position synchronous errors of the dual linear stages. Two velocity controllers and a synchronous controller are designed for the control system. To decrease the velocity tracking errors, two velocity controllers are designed to follow the velocity tracking commands which are also the desired velocity profiles specified by the given tasks and have optimized acceleration and deceleration characteristics. Moreover, the synchronous controller is designed to quickly and stably eliminate the position and velocity synchronous errors of the two dual linear stages by improving the damping characteristics of the synchronous control system. The experimental results are presented to demonstrate the effectiveness of the proposed synchronous control system.

In this study, the authors propose an adaptive scheme to improve motion estimation of a variational model based on image features and flow quality measurements. Using image features, the authors introduce adaptive functions and inject them into the energy function to fine‐tune the estimation process. They propose a hybrid scheme to deal with large motions and improve the accuracy of the flow field. They introduce a trusted‐map based on constraints to measure flow quality. They use this map as a reference for the proposed occlusion‐aware filter. The proposed filter and hybrid scheme are integrated to correct the flow field iteratively, thus significantly improving the estimation results. The filter also enhances the flow field in occlusion areas. The authors experimental results demonstrate that their method provides sharp flow fields and significantly improved estimation accuracy.

Binocular stereo matching is one of the most important algorithms in the field of computer vision. Adaptive support-weight approaches, the current state-of-the-art local methods, produce results comparable to those generated by global methods. However, excessive time consumption is the main problem of these algorithms since the computational complexity is proportionally related to the support window size. In this paper, we present a novel cost aggregation method inspired by domain transformation, a recently proposed dimensionality reduction technique. This transformation enables the aggregation of 2-D cost data to be performed using a sequence of 1-D filters, which lowers computation and memory costs compared to conventional 2-D filters. Experiments show that the proposed method outperforms the state-of-the-art local methods in terms of computational performance, since its computational complexity is independent of the input parameters. Furthermore, according to the experimental results with the Middlebury dataset and real-world images, our algorithm is currently one of the most accurate and efficient local algorithms.

High dynamic range conditions are major obstacles to the implementation of practical stereovision systems in real scenes. We address this problem by introducing an adaptive local ternary-derivative pattern (ALTDP) which is a fusion of the local ternary pattern (LTP) and local derivative pattern (LDP). We make three main contributions in this study: (i) ALTDP encodes more detail information than LDP by extending to eight directions; (ii) ALDTP is better at discriminating and less sensitive to noise in uniform regions with three-value encoding (−1,0,1) without using a pre-defined threshold; and (iii) ALTDP significantly improves the performance of hierarchical belief propagation (BP) by substituting ALTDP data cost for the different intensity data cost. Moreover, our proposed method performs slightly better than LBP and LDP with three datasets: synthetic sequences (set 2) in the EISATS dataset, bright differences sequences (set 5) in the EISATS dataset, and the bumblebee xb3 dataset.

Guided image filtering has recently emerged as an effective technique for noise reduction and edge-preserving smoothing operation due to its appealing properties. It outperforms conventional bilateral filtering in a variety of applications in terms of both quality and computational cost. However, the combination of smoothing parameters (ε and Ωs) that delivers optimal results has not yet been reported, and the contrast of the filtered output is considerably reduced. In this paper, we present a restricted version of guided filtering that has better contrast-preserving characteristics, and use Stein's unbiased risk estimate (SURE) with an exhaustive search or a linear expansion of threshold (LET) to optimally tune the two above parameters. With SURE, the mean squared error (MSE) can be unbiasedly estimated without the requirement of the noise-free image. Experiments verified the accuracy of the SURE derivation and its effectiveness with respect to providing a better trade off between two interrelated objectives - noise reduction and edge-preservation.

Spatiotemporal stereo matching has attracted much interest over the last few years. The problem is to maintain the temporally consistent disparity of video sequences and to avoid the flickering-artifacts occurring in consecutive disparity maps. In this paper, we propose a constant time spatiotemporal local stereo matching based on the information permeability method that was recently proposed and achieves good stereo results for static image pairs. The proposed three-pass aggregation method takes into account multiple preceding and following frames of the video instead of a single image pair as in the conventional method. Consequently, the temporal disparity consistency is enforced without requiring an explicit motion estimation as several spatiotemporal approaches have done. Experiments showed that our method improves the disparity consistency compared to the conventional information permeability method, and that it outperforms existing spatiotemporal stereo matching techniques.

A new approach for vehicle detection and distance estimation based on stereo vision and evolutionary algorithm (SEA) is described in this paper. First, we reuse our recent work on FPGA implementation of census-based correlations for stereo matching. Next, the SEA uses the gray scale left image and disparity information obtained from the FPGA system to detect the preceding vehicle and estimate its distance. This paper introduces an effective fitness function that allows our proposed method to have an improved performance and higher accuracy when compared with the existing evolutionary algorithm (EA) based methods. A new crossover type, tourna-ment crossover, is introduced to reduce the convergence time of our proposed. This paper also introduces a new approach for estimating the fitness function parameters. This estimation differs from the traditional EA because these parameters were generally created via experiments. Moreover, the processing time and accuracy of SEA can be improved by converting the global search to the local search with V disparity map. The robust experiments have proved that SEA successfully detects vehicles in front and sustains noise from different objects appearing along the road. The detection range is 10m-140m, the detection rate is 95% and the average processing-time is approximately 31 ms/frame on CPU. These results prove that SEA is suitable for a real-time system.

This article describes a reliable gateway for in-vehicle networks. Such networks include local interconnect networks, controller area networks, and FlexRay. There is some latency when transferring a message from one node (source) to another node (destination). A high probability of error exists due to different protocol specifications such as baud-rate, and message frame format. Therefore, deploying a reliable gateway is a challenge to the automotive industry. We propose a reliable gateway based on the OSEK/VDX components for in-vehicle networks. We also examine the gateway system developed, and then we evaluate the performance of our proposed system.

This paper presents design and implementation of a pipelined datapath for real-time face detection using cascades of boosted classifiers. We propose following methods: symmetric image downscaling, classifier sharing, and cascade merging, to achieve the desired processing speed and area efficiency. First, an image pyramid with 16 levels is generated from the input image to simultaneously detect faces with different scales. The downscaled images are then transferred to the first stage of the cascade that is shared between the corresponding image pairs based on the pixel validity of the symmetric image pyramid. The last method exploits the different hit ratios of the cascade stages. We use a tree-structured cascade of classifiers since most of the nonface elements are eliminated during the early stages of the classifier. The use of a synthesis tool confirms that the proposed design reduces resource utilization by one-eighth without accuracy loss, compared to the fully parallelized implementation of the same algorithm. We implemented the proposed hardware architecture on a Xilinx Virtex-5 LX330 FPGA. The indicative throughput is 307 frames/s irrespective of the number of faces in the scene for standard VGA (640 × 480) images with an operating frequency of 125.59 MHz. We may ensure that face detection results are generated at each clock cycle after the initial pipeline delay, using this fully pipelined datapath for tree-structured cascade classifiers.

This paper presents a variational model to compute the optical flow using image-driven functions. The intensity, gradient and smoothness have different influences on each image area. Thus, we propose the control functions that take the image as the input to tune the estimation process. We use the second moment matrix to characterize distinct image areas and embed these functions into the variational model. We also separate the gradient term and intensity term in the model. In addition, we use the coarse-to-fine strategy to deal with the large displacement in the image sequence. Experimental results show the stability of our proposed method on different image sequences.

Optical flow is a motion field estimation method that has a wide range of applications. In this paper, we present a fully pipelined hardware architecture for high-speed optical flow estimation based on a full-search block matching algorithm. A census transform is applied to the corresponding pixels in the current and previous frame. The similarity between two census vectors within the search area is then computed by measuring the hamming distance. Macro blocks are generated based on the measured hamming distance values and the best match is determined by locating the block that has the smallest sum. The synthesis tool reported that the proposed system is capable of processing 400 standard VGA frames per second.

Recent advances made in IT technology has given much impetus to the development of multimedia devices. The digital camera is such a multimedia device. It has made much progress and become very popular, as most people now own a digital camera or cell phone with camera features. People often take photographs in everyday life. Professional photographers often take photographs of travel destinations, banquet halls or parties. In this paper, we propose an autonomous robot photographer capable of taking pictures and thus replacing photographers. This photographer robot can detect direction based on the human voice. It can control composition based on skin color detection to snap the picture.

The use of edge-related applications is important in the field of computer vision. These applications help robots with understanding their surrounding environments; the lane or wall detection system is one of the most popular applications. Numerous studies have recently been conducted for enhancing the capabilities of robotic vision, but they typically lacked the applications that were related to coping with the environmental changes of the scenes. In this paper, we propose a method that integrates a particle filter into the process of tracking the camera's parameters (the exposure and the focus) to find the captured frame with the high edge quality. The relationship between the current sequence of frames and the previous sequence was given no consideration when all the possible parameters were scanned. Our work attempts to find that relationship and to increase the speed of the camera system. The edge results are evaluated with using a line detection algorithm - that is known as the Standard Hough Transform. A test method is applied to analyze the correctness of the line detection results. Furthermore, we propose the entropy of the Sobel gradient method for measuring the image sharpness and its contrast when the exposure and focus of a digital camera are changed. Our experimental results show that our method can be applied in real-time systems because of its low computational requirements.

Optical flow is an important problem in computer vision since applications of accurate optical flow estimation enable us to control and manipulate tracking, 3-D reconstruction, motion blurring, and dirt removal. Many powerful methods have been proposed to solve the optical flow problem; however, instabilities at the boundaries of moving objects are still challenges. A difficult part of the optical flow problem is how to accurately and quickly detect and readjust unstable regions at the boundaries. This paper aims to enhance optical flow estimation by detecting and readjusting the unstable regions. In this paper, a new algorithm to detect and quickly readjust unstable regions at the boundaries of moving objects is presented in a more general and compact manner. In addition, a new context-based anisotropic diffusion filter, which is significant in processing intermediate data, is discussed in detail. Our approach has demonstrated more accurate results than previous approaches.

Stereo vision is a well-known ranging method because it resembles the basic mechanism of the human eye. However, the computational complexity and large amount of data access make real-time processing of stereo vision challenging because of the inherent instruction cycle delay within conventional computers. In order to solve this problem, the past 20 years of research have focused on the use of dedicated hardware architecture for stereo vision. This paper proposes a fully pipelined stereo vision system providing a dense disparity image with additional sub-pixel accuracy in real-time. The entire stereo vision process, such as rectification, stereo matching, and post-processing, is realized using a single field programmable gate array (FPGA) without the necessity of any external devices. The hardware implementation is more than 230 times faster when compared to a software program operating on a conventional computer, and shows stronger performance over previous hardware-related studies.

This paper presents an FPGA-based parallel hardware architecture for real-time face detection. An image pyramid with twenty depth levels is generated using the input image. For these scaled-down images, a local binary pattern transform and feature evaluation are performed in parallel by using the proposed block RAM-based window processing architecture. By sharing the feature look-up tables between two corresponding scaled-down images, we can reduce the use of routing resources by half. For prototyping and evaluation purposes, the hardware architecture was integrated into a Virtex-5 FPGA. The experimental result shows around 300 frames per second speed performance for processing standard VGA (640times480times8) images. In addition, the throughput of the implementation can be adjusted in proportion to the frame rate of the camera, by synchronizing each individual module with the pixel sampling clock.

Skin color is widely used in many applications because of its merit in human-machine interactions. However, detecting skin color requires repetitive operations on all pixels in the image, similar to other vision-based applications. Since the per-pixel processing is difficult to perform efficiently in conventional computers, many real-time image processing applications have problems with performance. In this paper, we propose FPGA implementation of a real-time skin color detection system. Among the various skin color detection methods, we chose a parametric skin distribution modeling method based on a Gaussian mixture, due to its acceptable training amount and skin detection performance. In addition, a mean-based surface flattening method was also proposed and implemented to improve the detection performance. The proposed method flattens the surface of objects in the scene by replacing the pixel value with the mean of its similar neighborhoods to remove the color noise. After this flattening process, the pixel values of the analogous adjacent pixels are located within a narrow range and are easily segmented to a different region. To consider the inherent parallelism of local image processing, all these functions are implemented within the FPGA to meet the demands of real-time performance.

We present a method to detect human fingertips from images captured by a stereo camera. The system makes use of the disparity information from a stereo camera to find candidates, and defines an evaluation process to detect two hands. The finger detector then processes each hand image to extract finger images. Finally, we perform geometric calculations on the results to relocate the positions of the fingertips. The proposed method is not complex; however, it shows exciting results in terms of run time and detection rates. The extraction result can be used in hand configuration modeling for gesture recognition in HCI systems.

In this paper, we propose a motion-based game control system using the flex sensor and the infrared sensor board. This system operates using a combination of player movement data detected from the flex sensor and the direction or positional data detected from the infrared sensor board. The player's motion is detected from a flex sensor band worn on his/her elbow and knee and the flex sensor belt on his/her waist. The infrared sensor board detects the player's direction of movement through the detected sequence or detected position of the infrared sensors. This system does not recognize all the detailed human motion, since detailed motion detection is not necessary to control the game. The proposed system only detects player motion types through multiple sensors. This feature distinguishes the system from other motion-based controllers such as Nintendo Wii, Sony motion controller and Microsoft Natal. Unlike our system, these existing controllers require a high-technology, high-cost system. This system is implemented on an existing game pad (PC, Xbox360, PS3) based combat game to test the detection of the player motion type.

This research aim is to create sensible interactions between multisensors in mobile devices. The proposal covers sensible interactions for all three kinds of methods: Mobile Application Interaction, Mobile to Mobile Interaction, and Mobile to Home Appliance Interaction. Five components are built in the handheld sensible interface device: distance measurement component, acceleration measure component, main processing component, haptic generation component, and wireless communication component. This research can support emotional user experience better than the button type input method or touch type input method can do. In addition, the proposal developed from the research can support Mobile to Mobile Interaction and Mobile to Home Appliance Interaction. The implement of the proposed sensible interaction method will demonstrate an appliance with a navigation application and digital TV control.

This paper proposes a method for detecting the camera motion between two successive images to compensate background motion. The camera motion is restricted with regard to panning, tilting, and zooming. For small panning and tilting angles and small values of zooming difference, we assume that the apparent background motion occurring between two consecutive images can be approximated. We perform this using a scaling transformation followed by a translation. The vertical and horizontal histograms of two successive images are created and then matched using Hough transformations. The transformation parameters are determined when the vertical and horizontal histograms are matched. A multi-resolution Hough transformation is employed to reduce processing time.

Position sensors using magnetic effects, such as magnetic encoders (MEs), are increasingly used in many industrial applications. These include motor control, electro-mechanical braking systems, and precision measurement systems. The magnetic encoders generally provide a pair of sinusoidal signals dephasing 90° to each other. Unfortunately, the signals obtained are not ideal, in that they always exist with the DC offsets, different amplitudes, phase-shifts and waveform distortions. These are further compromised by noise and changes in operating conditions. This paper proposes an approach for real-time correcting and tracking ME signals based on the Advanced Adaptive Digital Phase-Locked Loop (AADPLL) technique. AADPLL provides a robust filtering characteristic as well as a wideband of input frequency. It can effectively filter noise and improve the accuracy of ME signals. AADPLL also takes advantage of tracking high-speed input signals without time-lag, unlike traditional filters. A quadrature pulse interpolator is introduced to obtain high resolution. This interpolator uses look-up tables (LUTs) to store pulse-out values in one period of MEs signals. These LUTs are reduced in size for easy implementation in Digital Signal Processor (DSP) hardware platforms. Experimental results demonstrate the effectiveness of the proposed method.

Object tracking is a challenging problem in a number of computer vision applications. A number of approaches have been proposed and implemented to track moving objects in image sequences. The particle filter, which recursively constructs the posterior probability distributions of the state space, is the most popular approach. In the particle filter, many kinds of features are used for tracking a moving object in cluttered environments. The specific feature for tracking is selected according to the type of moving object and condition of the tracking environment. Improved tracking performance is obtained by using multiple features concurrently. This paper proposes the particle filter algorithm, using multiple features, such as IFD (inter-frame difference) and gray level, to track a moving object. The IFD is used to detect an object and the gray level is used to distinguish the target object from other objects. This paper designs the circuit of the proposed algorithm using VHDL (VHSIC hardware description language) in an FPGA (field programmable gate array) for tracking without considerable computational cost, since the particle filter requests many computing powers to track objects in real-time. All functions of the proposed tracking system are implemented in an FPGA. A tracking system with this FPGA is implemented and the corresponding performance is measured

A variety of devices are used for robot control such as personal computers or other human interface devices, haptic devices, and so on. However, sometimes it is not easy to select a device which fits the specific character of varied kinds of robots while at the same time increasing the user's convenience. Under these circumstances, in this study, we have tried to measure user convenience. We tried to understand the characteristics of several devices used to achieve human robot interaction by using each of these devices that could be used with a personal computer: We used a button type device, a joystick, a driving device which consisted of a handle and pedals, and a motion-based human interface device including an acceleration sensor.

A robot highly motivates students and it is one of the best ways to connect students with the technology. LEGO created a set called robotic invention system. This system helps students to understand the technology of both robot and programming language. It also improves creativeness by building and controlling the robot. This paper would propose the idea on educating C language to students using Robotic Invention System 2.0

Particle filters have attracted much attention due to their robust tracking performance in cluttered environments. Particle filters maintain multiple hypotheses simultaneously and use a probabilistic motion model to predict the position of the moving object, and this constitutes a bottleneck to the use of particle filtering in real-time systems due to the expensive computations required. In order to track moving objects in real-time without delay and loss of image sequences, a particle filter algorithm specifically designed for a circuit and the circuit of the object tracking algorithm using the particle filter are proposed. This circuit is designed by VHDL (VHSIC hardware description language), and implemented in an FPGA (field programmable gate array). All of the functions of the proposed particle filter used to track moving objects are implemented in the FPGA. The object tracking system employing this circuit is implemented and then its performance is measured

Tactile sensation is one of the most important sensory functions along with the auditory sensation for the visually impaired because it replaces the visual sensation of the persons with sight. In this paper, we present a tactile display device as a dynamic Braille display that is the unique tool for exchanging information among them. The proposed tactile cell of the Braille display is based on the dielectric elastomer and it has advantageous features over the existing ones with respect to intrinsic softness, ease of fabrication, cost effectiveness and miniaturization. We introduce a new idea for actuation and describe the actuating mechanism of the Braille pin in details capable of realizing the enhanced spatial density of the tactile cells. Finally, results of psychophysical experiments are given and its effectiveness is confirmed.

In this paper we present a packaged actuator to be applied for micro- and macro- robotic applications. The actuator is based on polymer dielectrics, and intrinsically has muscle-like characteristics capable of performing motions such as forward/backward/controllable compliance. The actuator is featured in several aspects such as simplicity and lightness in weight, cost-effectiveness, multiple DOF-actuation, and digital interface. In this paper, its basic concepts are briefly introduced and the issues about design, fabrication and applications are discussed.

In order to make electronic products such as personal digital assistants (PDA) and cellular phones be small and thin, their components must be also small and thin. Since it is very difficult to inspect these small components by human eyes, an automation system for inspecting them has been used. Compared with inspection systems for large or medium size components, those for small components need to be more precise. In particular, more precise timing for each action in the inspection system of small components is required. Programmable logic controllers (PLC) have been used for generating precise timing signals. However, it is neither convenient nor economical to generate timing signals by PLC. In this paper, a low-cost programmable timing controller to generate precise signals in the inspection system of small components is proposed and it is applied to an inspection system of multilayer chip capacitors (MLCC). Since the proposed controller is organized to have internal registers, counters, and software routines for generating timing signals, users do not have to program the details about timing signals and need not only send some values about an inspection system through an RS232C port. By selecting these values appropriate for a given inspection system, desired timing signals could be generated.

In this paper a new soft actuator based on dielectric elastomer is proposed. The actuator, called an antagonistically-driven linear actuator (ANTLA), has the muscle-like characteristics capable of performing the motions such as forward/backward/controllable compliance. Due to its simplicity of configuration and ease of fabrication, it has the advantage to be scale-independently implemented in meso- or micro-scale robotic applications. Its basic concepts are introduced and a quasi-static analysis is performed with experimental verifications.

In this paper a new soft actuator based on dielectric elastomer is proposed. The actuator, called an antagonistically-driven linear actuator (ANTLA), has the muscle-like characteristics capable of performing the motions such as forward/backward/controllable compliance. In this paper, its dynamic analysis is performed with experimental verifications, and applications for robotic actuating devices are introduced.

In this paper a microrobot, mimicking annelid animals like the earthworm, is presented. The robot is composed of several ring-like segments. Each segment is able to generate three degree-of-freedom motions such as pan/tilt/up-down respectively, and it is actuated by three soft actuators located equidistantly along the circumferential direction. The soft actuator, called antagonistically-driven linear actuator (ANTLA), is based on polymer dielectrics and has muscle-like characteristics capable of performing motions such as forward/backward/controllable compliance. In this paper, the basic concept of the actuator is briefly reviewed and the robot, including the ring-like segment, is explained in detail with demonstrations.

Ever since the first power plant had been built in 1977, the electricity by nuclear power plants has been increased in Korea. It has been required that the spent nuclear fuels of nuclear power plants should be managed safely and therefore the technology about this should be also developed. Korea Atomic Energy Research Institute has developed the devices to manage these spent nuclear fuels. Due to high radioactivity, all these devices should be operated in a hot cell, which is a sealed room. Since these devices should be very high reliable, a real-time monitoring is necessary to check that they are working correctly. In this paper, a real-time 3 dimensional graphic simulator is proposed for the monitoring and control of the spent nuclear fuel dismantlement robots through the Internet. In order to reduce the visualization time of the devices, the abstraction of graphics data is performed. Also, simple operation information from large sensor information is extracted and an efficient message format and its communication scheme are defined to reduce the communication time over the Internet.

If we install an electrode such as conductive grease or carbon powder in both sides of an electrostrictive polymer film, then it becomes a parallel-plate capacitor using electrostrictive polymer as dielectric. When the voltage is applied to electrodes, charges are piled up on the electrodes. Then, there is an attractive force between charges in the anode and cathode. This force acts as pressure in the direction of thickness, so electrostrictive polymer expands in the direction of area by its incompressible characteristic. When the voltage is off, electrostrictive polymer returns to its original shape. In this paper, we realize an electrostrictive polymer actuator using the deformation of electrostrictive polymer, and design and implement the control system of this actuator. Using this control system, we analyzed its actuation and motion control. In order to do that, we analyzed the relations between voltage and displacement. Also, the characteristics of this actuator were measured under some control techniques.