To address the unique attributes of three-dimensional point cloud data, this paper introduces an innovative architecture for precise 3D point cloud classification. Expanding on the PointMLP framework, we incorporate an embedding module that elevates the point cloud to higher-dimensional feature representations which are followed by geometric feature mapping and extraction modules to capture point cloud characteristics. To estimate local geometric structures, we use plane features to determine planes associated with nearby points. Additionally, we integrate self-attention mechanisms to capture intricate local geometric features. Moreover, MLP modules with residual connections are employed for efficient feature extraction. The derived features are then reduced in size using Max Pooling layers. For classification purposes, we utilize fully connected layers, batch normalization, activation functions, and random weight dropping techniques to enhance generalization and ensure robustness on unseen data. By adopting these architectural decisions, our proposed model achieves significant progress in accurately classifying 3D point clouds.

Sample-efficient neural architecture search (NAS) techniques have advanced rapidly. Two lines of methods, namely neural predictor and sequential search, have shown promising performance in improving the sample efficiency of NAS. However, as far as we know, little attention has been paid to the middle ground between these two lines. Inspired by the analogy between NAS and evolutionary optimization, we propose a new Sample-Efficient Training for NAS (SETNAS) based on strategies that improve fitness scores and sampling mechanisms. We develop a strong neural predictor called the Fully Bidirectional Graph Convolutional Network evolutionary (Fully-BiGCN) that significantly enhances the predictor capability of the features in each layer. The developed predictor is embedded into an iterative stratified sampling process to retain only a subset of best-fit architectures using the same training budget. SET-NAS achieves remarkable results compared to the state-of-the-art in predictor-based NAS. Using NASBench-201 as the benchmark, SET-NAS takes only $27.1 \%$ (CIFAR-10), $49.0 \%$ (CIFAR-100), and $51.75 \%$ (ImageNet-16) of training cost of other state-of-the-art predictor-based methods to find the promising network architecture.

Neural Architecture Search (NAS) methods seek effective optimization toward performance metrics regarding model accuracy and generalization while facing challenges regarding search costs and GPU resources. Recent Neural Tangent Kernel (NTK) NAS methods achieve remarkable search efficiency based on a training-free model estimate; however, they overlook the non-convex nature of the DNNs in the search process. In this paper, we develop Multi-Objective Training-based Estimate (MOTE) for efficient NAS, retaining search effectiveness and achieving the new state-of-the-art in the accuracy and cost trade-off. To improve NTK and inspired by the Training Speed Estimation (TSE) method, MOTE is designed to model the actual performance of DNNs from macro to micro perspective by draw loss landscape and convergence speed simultaneously. Using two reduction strategies, the MOTE is generated based on a reduced architecture and a reduced dataset. Inspired by evolutionary search, our iterative ranking-based, coarse-to-fine architecture search is highly effective. Experiments on NASBench-201 show MOTE-NAS achieves 94.32% accuracy on CIFAR-10, 72.81% on CIFAR-100, and 46.38% on ImageNet-16-120, outperforming NTK-based NAS approaches. An evaluation-free (EF) version of MOTE-NAS delivers high efficiency in only 5 minutes, delivering a model more accurate than KNAS.

Mangrove canopy height information is crucial to carbon stock and biomass analyses. However, estimation of this height is challenging because of large areas involved, and field conditions of mangrove forests. Remote sensing satellite imagery has been used for canopy height mapping because it offers several advantages. This study developed a spatial–spectral–temporal deep learning regression model with convolutional long short-term memory (ConvLSTM) and transformer (hereafter referred to as the SST-CLT model) to map mangrove canopy height over large area. The SST-CLT model consists of two sub-models trained simultaneously. The first sub-model is fusion extractor to extracts spatial–spectral–temporal information from Sentinel-1 time-series data by using a ConvLSTM. It also extracts spatial–spectral information from Sentinel-2 data using a two-dimensional convolutional block. The second sub-model is a regressor contains Swin transformer and final convolutional regression layer. Data from light detection and ranging canopy height model were employed as the target data to train the proposed model. The SST-CLT model was tested on two datasets collected from Florida: large dataset for the Everglades National Park (ENP) and small dataset for the Charlotte Harbor Preserve State Park (CHPSP). The SST-CLT model achieved a mean absolute error (MAE) of 1.924 and 1.913 m for the ENP and CHPSP datasets, respectively. Moreover, it achieved root mean square error (RMSE) values of 2.471 and 2.440 m for these datasets, respectively. The SST-CLT model was compared with that of other regression models. The results indicated that the MAE and RMSE of the proposed SST-CLT were lower than those of the other models. https://github.com/ilhamjamal/SST-CLT.

FPN (Feature Pyramid Network) has become a basic component of most SoTA one stage object detectors. Many previous studies have repeatedly proved that FPN can caputre better multi-scale feature maps to more precisely describe objects if they are with different sizes. However, for most backbones such VGG, ResNet, or DenseNet, the feature maps at each layer are downsized to their quarters due to the pooling operation or convolutions with stride 2. The gap of downscaling-by-2 is large and makes its FPN not fuse the features smoothly. This paper proposes a new SFPN (Synthetic Fusion Pyramid Network) arichtecture which creates various synthetic layers between layers of the original FPN to enhance the accuracy of light-weight CNN backones to extract objects’ visual features more accurately. Finally, experiments prove the SFPN architecture outperforms either the large backbone VGG16, ResNet50 or light-weight backbones such as MobilenetV2 based on AP score.

The development of lightweight object detectors is essential due to the limited computation resources. To reduce the computation cost, how to generate features plays a significant role. This paper proposes a new lightweight convolution method Cross-Stage Lightweight Module (CSL-M). It combines the Inverted Residual Block (IRB) and Cross-Stage Partial (CSP) concept. Experiments conducted at CIFAR-10 show that the proposed CSL-Net based on CSL-M performs better with fewer FLOPs than the other lightweight backbones. Finally, we use CSL-Net as the backbone to construct a lightweight detector CSL-YOLO, achieving better detection performance with only 43% FLOPs and 52% parameters than Tiny-YOLOv4.

In inverse synthetic aperture radar (ISAR) imaging, the imaging results can be affected by the unexpected target motions. This results in a blurry and unrecognizable image. The motion parameters estimation is a compensation method to improve the ISAR image refocusing and quality. In this study, the backscattered echo signals are simulated by the linear geometry system of ISAR moving targets. The entropy of ISAR image is used as a criterion to evaluate the image quality. Furthermore, this entropy measure can be treated as a cost function of the particle swarm optimization (PSO) method and minimized by PSO to improve the quality of ISAR images. The experimental results showed that our proposed PSO motion estimation approach to entropy minimization for ISAR imaging can not only efficiently improve the estimation capability of motion parameters, but also significantly achieve a better performance of ISAR image refocusing.

Investigating the identity, distribution, and evolution of bird species is important for both biodiversity assessment and environmental conservation. The discrete wavelet transform (DWT) has been widely exploited to extract time-frequency features for acoustic signal analysis. Traditional approaches usually compute statistical measures (e.g., maximum, mean, standard deviation) of the DWT coefficients in each subband independently to yield the feature descriptor, without considering the intersubband correlation. A new acoustic descriptor, called the local wavelet acoustic pattern (LWAP), is proposed to characterize the correlation of the DWT coefficients in different subbands for birdsong recognition. First, we divide a variable-length birdsong segment into a number of fixed-duration texture windows. For each texture window, several LWAP descriptors are extracted. The vector of locally aggregated descriptors (VLAD) is then used to aggregate the set of LWAP descriptors into a single VLAD vector. Finally, principal component analysis (PCA) plus linear discriminant analysis (LDA) are employed to reduce the feature dimensionality for classification purposes. Experiments on two birdsong datasets show that the proposed LWAP descriptor outperforms other local descriptors, including linear predictive coding cepstral coefficients, Mel-frequency cepstral coefficients, perceptual linear prediction cepstral coefficients, chroma features, and prosody features. Furthermore, the proposed LWAP descriptor, followed by VLAD encoding, PCA plus LDA feature extraction, and a simple distance-based classifier, yields promising results that are competitive with those obtained by the state-of-the-art convolutional neural networks.

The focus of content-based image retrieval (CBIR) is to narrow down the gap between low-level image features and high-level semantic concepts. In this paper, a biased discriminant analysis with feature line embedding (FLE-BDA) is proposed for performance enhancement in relevance feedback schemes. Maximizing the margin between relevant and irrelevant samples at local neighborhoods was the aim in this study. In reduced subspace, relevant images and query images can be quite close, while irrelevant samples are far away from relevant samples. The results of four benchmark datasets are given to show the performance of the proposed method.

Due to the development of economy and technology, the people's demands on cars are growing and so are the problems, such as finding stolen car, banning violation and parking lot management. It will be time-consuming and low-efficiency if we do those jobs only by human because of the limitation of human being's concentration. Therefore, it has been a popular topic to develop intelligent monitoring system under new video technology within a decade. There are still many rooms for future development and application in the car license detection and recognition fields. For finding stolen car, we can integrate car license detection system and road monitoring system to analyze the videos and trace the objects, so we can gain high-efficiency and low-cost results. For parking lot management system, we can have the result of access management and automatic charge to reduce the human resource through car license recognition system. Automated toll of highway can be done through car license recognition system as well. Because the car license recognition system is mostly applied to security monitoring field and business purposes, the demand of the accuracy is quite strict. There are causes which make the inaccuracy of the license recognition, such as the lack of video resolution, too small license due to the distance and the light and shadow. We will discuss the license recognition system and how to use apply the super-resolution to overcome those above problems.

In this paper, a novel local pattern descriptor generated by the proposed local vector pattern (LVP) in high-order derivative space is presented for face recognition. The proposed vector representation of the referenced pixel is generated to provide the one-dimensional structure of micropatterns. To effectively extract more detailed discriminative information in a given sub-region, the vector of LVP is refined by varying local derivative directions from the nth-order LVP in (n-1)th-order derivative space. The proposed LVP is compared with the existing local pattern descriptors including local binary pattern (LBP), local derivative pattern (LDP), and local tetra pattern (LTrP) to evaluate the performances from input grayscale face images. Extensive experiments conducting on benchmark face image databases, FERET and Extended Yale B, demonstrate that the proposed LVP in high-order derivative space indeed performs much better than LBP, LDP and LTrP for face recognition.

In this paper, a novel local pattern descriptor generated by the proposed local vector pattern (LVP) in high-order derivative space is presented for use in face recognition. Based on the vector of each pixel constructed by computing the values between the referenced pixel and the adjacent pixels with diverse distances from different directions, the vector representation of the referenced pixel is generated to provide the 1D structure of micropatterns. With the devise of pairwise direction of vector for each pixel, the LVP reduces the feature length via comparative space transform to encode various spatial surrounding relationships between the referenced pixel and its neighborhood pixels. Besides, the concatenation of LVPs is compacted to produce more distinctive features. To effectively extract more detailed discriminative information in a given subregion, the vector of LVP is refined by varying local derivative directions from the n th-order LVP in (n-1) th-order derivative space, which is a much more resilient structure of micropatterns than standard local pattern descriptors. The proposed LVP is compared with the existing local pattern descriptors including local binary pattern (LBP), local derivative pattern (LDP), and local tetra pattern (LTrP) to evaluate the performances from input grayscale face images. In addition, extensive experiments conducting on benchmark face image databases, FERET, CAS-PEAL, CMU-PIE, Extended Yale B, and LFW, demonstrate that the proposed LVP in high-order derivative space indeed performs much better than LBP, LDP, and LTrP in face recognition.

This paper presents a reliable and robust palmprint verification approach that involves using a bi-feature, biometric, palmprint feature-point number (FPN) and a histogram of oriented gradient (HOG). The bi-feature was fused and verified using a support vector machine (SVM) at the feature level. The approach has the advantages of capturing palm images in pegless scenarios with a low cost and low-resolution (100 dpi) digital scanner, and one sensor can capture palmprint bi-feature information. The low-resolution images result in a smaller database. Nine thousand palmprint images were collected from 300 people to verify the validity of the proposed approach. The results showed an accurate classification rate of 99.04%. The experimental results demonstrated that the proposed approach is feasible and effective in palmprint verification. Our findings will help extend palmprint verification technology to security access control systems.

Face recognition algorithms often have to solve problems such as facial pose, illumination, and expression (PIE). To reduce the impacts, many researchers have been trying to find the best discriminant transformation in eigenspaces, either linear or nonlinear, to obtain better recognition results. Various researchers have also designed novel matching algorithms to reduce the PIE effects. In this study, a nearest feature space embedding (called NFS embedding) algorithm is proposed for face recognition. The distance between a point and the nearest feature line (NFL) or the NFS is embedded in the transformation through the discriminant analysis. Three factors, including class separability, neighborhood structure preservation, and NFS measurement, were considered to find the most effective and discriminating transformation in eigenspaces. The proposed method was evaluated by several benchmark databases and compared with several state-of-the-art algorithms. According to the compared results, the proposed method outperformed the other algorithms.

This paper presents a novel segmentation algorithm to segment a body posture into different body parts using the technique of deformable triangulation. To analyze each posture more accurately, they are segmented into triangular meshes, where a spanning tree can be found from the meshes using a depth-first search scheme. Then, we can decompose the tree into different subsegments, where each subsegment can be considered as a limb. Then, two hybrid methods (i.e., the skeleton-based and model-driven methods) are proposed for segmenting the posture into different body parts according to its occlusion conditions. To analyze occlusion conditions, a novel clustering scheme is proposed to cluster the training samples into a set of key postures. Then, a model space can be used to classify and segment each posture. If the input posture belongs to the nonocclusion category, the skeleton-based method is used to divide it into different body parts that can be refined using a set of Gaussian mixture models (GMMs). For the occlusion case, we propose a model-driven technique to select a good reference model for guiding the process of body part segmentation. However, if two postures' contours are similar, there will be some ambiguity that can lead to failure during the model selection process. Thus, this paper proposes a tree structure that uses a tracking technique so that the best model can be selected not only from the current frame but also from its previous frame. Then, a suitable GMM-based segmentation scheme can be used to finely segment a body posture into the different body parts. The experimental results show that the proposed method for body part segmentation is robust, accurate, and powerful.

Falsification of human faces using face photos has been an arising problem for face recognition and verification systems. In this paper, we propose a system to distinguish face photos from true faces by their motion models. In order to enhance the difference between the two classes, we design an enhanced optical flow method which generates a larger difference between the motion model of true faces and that of face photos. The feature vector we adopted is the dense optical flow field across a short period of time. An LDA-based training method is adopted to separate the projection of the training data into two classes, and a Bayes classifier is used to classify the testing samples. Under the specified motion of true faces and face photos, our proposed method can effectively distinguish the two classes with high verification rate. Even if the motion is arbitrary for both classes, the proposed system can also report satisfying results.

This letter proposes a novel method to detect carried objects from videos and applies it for analysis of suspicious events. First of all, we propose a novel kernel-based tracking method for tracking each foreground object and further obtaining its trajectory. With the trajectory, a novel ratio histogram is then proposed for analyzing the interactions between the carried object and its owner. After color re-projection, different carried objects can be then accurately segmented from the background by taking advantages of Gaussian mixture models. After bag detection, an event analyzer is then designed to analyze various suspicious events from the videos. Even though there is no prior knowledge about the bag (such as shape or color), our proposed method still performs well to detect these suspicious events. As we know, due to the uncertainties of the shape and color of the bag, there is no automatic system that can analyze various suspicious events involving bags (such as robbery) without using any manual effort. However, by taking advantages of our proposed ratio histogram, different carried bags can be well segmented from videos and applied for event analysis. Experimental results have proved that the proposed method is robust, accurate, and powerful in carried object detection and suspicious event analysis.

This paper presents a hierarchical lane detection system with the ability to deal with both structured and unstructured roads. The proposed system classifies the environment first before applying suitable algorithms for different types of roads. For environment classification, pixels with lane- marking colors are extracted as feature points. Eigenvalue decomposition regularized discriminant analysis is utilized in model selection and maximum likelihood estimation of Gaussian parameters in high dimensional feature space. For structured roads, the extracted feature points are reused for lane detection. For unstructured roads, mean-shift segmentation is applied to divide the scene into regions. Possible road boundary candidates are selected, and Bayes rule is used to choose the most probable boundary pairs. The experimental results show that the system is able to robustly find the boundaries of the lanes on different types of roads and various weather conditions.

This paper proposes a novel method to detect suspicious objects from videos for abnormal event analysis. When considering a robbery event happens, there should be some suspicious object transferring conditions following between the forager and the victim. Since there is no prior knowledge about the object’s property, it is difficult to automatically analyze the conditions without any manual efforts. To tackle this problem, a ratio histogram based on fuzzy c-means algorithm is proposed for finding suspicious objects. Furthermore, we use Gaussian mixture models to model the suspicious object’s visual properties so that it can be accurately segmented from videos. After analyzing its subsequent motion features, different abnormal events like robbery can be effectively detected from videos. Experiment results have proved that the proposed method is robust, accurate, and powerful in abnormal event detection.

This paper proposes a novel method to detect suspicious objects from videos for robbery event analysis. First of all, a background subtraction using a minimum filter is used for detecting foreground objects from videos. Then, a novel kernel-based tracking method is proposed for tracking each moving object and obtaining its trajectory. Then, we propose a novel robbery event analysis system to analyze suspicious object transferring conditions between any two persons. Usually, when a robbery event happens, there should some suspicious object transferring conditions happening between the robbery and the victim. Since there is no prior knowledge about the object's property, it is difficult to automatically analyze the conditions without any manual efforts. To tackle this problem, a novel ratio histogram is then proposed for finding suspicious objects and then accurately analyzing their transferring conditions. After color re-projection, we use Gaussian mixture models to model the suspicious object's visual properties so that it can be very accurately segmented from videos. After analyzing its subsequent speed, different robbery events can be then effectively detected from videos. Experiment results have proved that the proposed method is robust, accurate, and powerful in robbery event detection.

This paper presents a novel template matching method to detect and track pedestrians for people counting in real-time. Firstly, a novel background subtraction method is proposed for extracting all foreground objects from background. Then, a shadow elimination method is used to remove unwanted shadow from the background. In order to identify pedestrians from non-pedestrian objects, this paper proposed a novel grid-based template matching scheme to robustly verify each pedestrian. Usually, a pedestrian will have different appearances at different positions. The grid-based approach can effectively reduce the perspective effects into a minimum since it uses different templates to record the appearance changes at each grid. When more templates are used, the detection process will become more inefficient. To speed up its efficiency, an integral image is used to filter out all impossible candidates in advance. Lastly, a tracking method is applied to tracking the direction of each moving pedestrian so that the real number of passing people per direction can be counted more accurately. Experimental results have proved that the proposed method is robust, accurate, and powerful in people counting.

This paper presents a novel vehicle detection approach for detecting vehicles from static images using color and edges. Different from traditional methods, which use motion features to detect vehicles, this method introduces a new color transform model to find important "vehicle color" for quickly locating possible vehicle candidates. Since vehicles have various colors under different weather and lighting conditions, seldom works were proposed for the detection of vehicles using colors. The proposed new color transform model has excellent capabilities to identify vehicle pixels from background, even though the pixels are lighted under varying illuminations. After finding possible vehicle candidates, three important features, including corners, edge maps, and coefficients of wavelet transforms, are used for constructing a cascade multichannel classifier. According to this classifier, an effective scanning can be performed to verify all possible candidates quickly. The scanning process can be quickly achieved because most background pixels are eliminated in advance by the color feature. Experimental results show that the integration of global color features and local edge features is powerful in the detection of vehicles. The average accuracy rate of vehicle detection is 94.9%.

In this paper, we propose the use of motion vectors embedded in MPEG bitstreams to generate so-called ldquomotion flowsrdquo, which are applied to perform video retrieval. By using the motion vectors directly, we do not need to consider the shape of a moving object and its corresponding trajectory. Instead, we simply ldquolinkrdquo the local motion vectors across consecutive video frames to form motion flows, which are then recorded and stored in a video database. In the video retrieval phase, we propose a new matching strategy to execute the video retrieval task. Motions that do not belong to the mainstream motion flows are filtered out by our proposed algorithm. The retrieval process can be triggered by query-by-sketch or query-by-example. The experiment results show that our method is indeed superb in the video retrieval process.

A lane-detection method aimed at handling moving vehicles in the traffic scenes is proposed in this brief. First, lane marks are extracted based on color information. The extraction of lane-mark colors is designed in a way that is not affected by illumination changes and the proportion of space that vehicles on the road occupy. Next, for vehicles that have the same colors as the lane marks, we utilize size, shape, and motion information to distinguish them from the real lane marks. The mechanism effectively eliminates the influence of passing vehicles when performing lane detection. Finally, pixels in the extracted lane-mark mask are accumulated to find the boundary lines of the lane. The proposed algorithm is able to robustly find the left and right boundary lines of the lane and is not affected by the passing traffic. Experimental results show that the proposed method works well on marked roads in various lighting conditions

In this paper, a novel italic detection and rectification method without the prerequisite of character recognition is proposed. The merit of the proposed method is that it can be operated directly on a single character so that more accurate statistical information can be obtained. The rationale of our proposed method is that the difference of certain features derived from italic style characters after shear transformation will be canceled, whereas the difference will be more obvious for non-italic style (normal style) characters.

This paper presents a novel vehicle detection approach for detecting vehicles from static images using color and edges. Different from traditional methods which use motion features to detect vehicles, this method introduces a new color transform model to find important "vehicle color" from images for quickly locating possible vehicle candidates. Since vehicles have different colors under different lighting conditions, there were seldom works proposed for detecting vehicles using colors. This paper proves that the new color transform model has extreme abilities to identify vehicle pixels from backgrounds even though they are lighted under various illumination conditions. Each detected pixel corresponds to a possible vehicle candidate. Then, two important features including edge maps and coefficients of wavelet transform are used for constructing a multi-channel classifier to verify this candidate. According to this classifier, we can perform an effective scan to detect all desired vehicles from static images. Since the color feature is first used to filter out most background pixels, this scan can be extremely quickly achieved. Experimental results show that the integrated scheme is very powerful in detecting vehicles from static images. The average accuracy of vehicle detection is 94.5%.

Gradual shot change detection is one of the most important research issues in the field of video indexing/retrieval. Among the numerous types of gradual transitions, the dissolve-type gradual transition is considered the most common one, but it is also the most difficult one to detect. In most of the existing dissolve detection algorithms, the false/miss detection problem caused by motion is very serious. In this paper, we present a novel dissolve-type transition detection algorithm that can correctly distinguish dissolves from disturbance caused by motion. We carefully model a dissolve based on its nature and then use the model to filter out possible confusion caused by the effect of motion. Experimental results show that the proposed algorithm is indeed powerful.

One of the key challenges for a watermarking scheme to be mandated in a digital right management (DRM) system is the robustness. This paper is focused on exploring the robustness against the watermark-estimation attacks (WEAs) that are clever at disclosing hidden information for unauthorized purposes without sacrificing media's quality. In WEAs, the collusion attack naturally occurs in video watermarking while the copy attack adapts to any media watermarking. In view of this, the aim of this study is to deal with the WEAs by means of a video frame-dependent watermark (VFDW). We begin by gaining insight into the WEAs, leading to formal definitions of "optimal watermark prediction" and "perfect cover data recovery". Subject to these definitions, the video-frame hash is addressed as a constituent component of the VFDW for antiestimation of hidden watermarks. Both mathematical analyses and experiment results consistently verify the antidisclosure capability of the video content-dependent watermarking scheme. Our approach is the first work that takes resistance to both the collusion and copy attacks into consideration.

Due to the emergence of digital libraries and the fast development of multimedia, more and more people use data hiding techniques to hide annotations or side information in images. We propose a novel data hiding mechanism based on pair-wise logical computation. The proposed mechanism can achieve the benefits of reversible and lossless reconstruction of the hidden data and the host image without utilizing any information from the original host image. It does not degrade the visual quality of the recovered host image after extracting the hidden data. Moreover, satisfactory data hiding capacity can be obtained simultaneously. The proposed data hiding mechanism is suitable for applying to the data hiding of images, scanned texts, figures, and signatures, especially for side information and annotation data embedding

This paper proposes a significant motion vector protection (SMVP) scheme for error-resilient transmission of videos. In terms of a rate-distortion optimization model, we show how to determine the significant motion vectors (SMVs) and how much rate space should be preserved to store SMVs. The idea behind our method is to give more protection to significant data. As a result, our method can be regarded as a kind of unequal error protection mechanisms. In comparison with the conventional forward error correction (FEC) and error concealment methods, experimental results demonstrate the effectiveness of the proposed method. In particularly, our method shows its superiority under the situation of higher packet loss rates.

A novel approach to personal verification using the thermal images of palm-dorsa vein patterns is presented in this paper. The characteristics of the proposed method are that no prior knowledge about the objects is necessary and the parameters can be set automatically. In our work, an infrared (IR) camera is adopted as the input device to capture the thermal images of the palm-dorsa. In the proposed approach, two of the finger webs are automatically selected as the datum points to define the region of interest (ROI) on the thermal images. Within each ROI, feature points of the vein patterns (FPVPs) are extracted by modifying the basic tool of watershed transformation based on the properties of thermal images. According to the heat conduction law (the Fourier law), multiple features can be extracted from each FPVP for verification. Multiresolution representations of images with FPVPs are obtained using multiple multiresolution filters (MRFs) that extract the dominant points by filtering miscellaneous features for each FPVP. A hierarchical integrating function is then applied to integrate multiple features and multiresolution representations. The former is integrated by an inter-to-intra personal variation ratio and the latter is integrated by a positive Boolean function. We also introduce a logical and reasonable method to select a trained threshold for verification. Experiments were conducted using the thermal images of palm-dorsas and the results are satisfactory with an acceptable accuracy rate (FRR:2.3% and FAR:2.3%). The experimental results demonstrate that our proposed approach is valid and effective for vein-pattern verification.

This study attempts to apply pattern recognition (PR) technologies with associative memory to real-time pattern recognition of engineering components using a client-server network structure in a Web-based recognition system. A remote engineer is able to draw directly the shape of engineering components using the browser, and the recognition system will search for the component database of a company using the Internet. Component patterns are stored in the database system. Their properties and specifications are also attached to the data field of each component pattern except that of engineering components. In our approach, the recognition system adopts parallel computing, and will raise the recognition rate. Our recognition system is a client-server network structure using the Internet. The system uses a recurrent neural network (RNN) with associative memory to perform training and recognition. The last phase utilizes the technology of database matching and solves the problem of spurious state. Our system will be used at the Yang-Fen Automation Electrical Engineering Company.

This study attempts to apply the principle of neural networks and pattern recognition (PR) technologies to real-time recognition by client-server network structure into a web-based recognizing system. In this paper, we recommend a Web-Based PR technology, which is improved recurrent neural network (RNN) from possessing feedback and non-linear activation function with its input, be taken out threshold. The purpose of this article is. to construct a Client-Server network structure for PR system with associative memory. The Server-end is built a databases management system for storage sample patterns. In proceed with training, the user can real-time assign any pattern, which is a record in the Server-end databases. In deal with retrieve task, we propose a novel PR method via databases matching, it can efficient solve spurious states problem from RNN in the WBPR system. On the other hand, taking advantage of Database Matching is to overcome the capacity restrictions on RNN. In order to clarify and corroborate the above Web-Based PR technology, thus a simulation experiment will be presented and their algorithms are also discussed.

Form recognition is one of the special applications of document analysis (DA). We present a novel form recognition method by analyzing the line structure embedded in an input form document. First, all vertical and horizontal lines embedded in the form image are extracted. By analyzing the crossing relationships among horizontal lines and vertical lines, a line crossing relationship matrix can be built with each row corresponding to one horizontal line and each column corresponding to one vertical line. Moreover two line distance relationship matrices, horizontal and vertical line distance relationship matrices, are built by analyzing the distance relationships among horizontal lines and vertical lines, respectively. Last, the recognition task is performed by matching these three matrices. Experimental results reveal the feasibility and efficiency of our proposed method in recognizing form documents.

Marginal noise is a common phenomenon in document analysis which results from the scanning of thick documents or skew documents. It usually appears in the front of a large and dark region around the margin of document images. Marginal noise might cover meaningful document objects. The overlapping of marginal noise with meaningful objects makes it difficult to perform the task of segmentation and recognition of document objects. This paper proposes a novel approach to remove marginal noise. The proposed approach consists of two steps which are marginal noise detection and marginal noise deletion. Marginal noise detection will reduce an original document image into a smaller image, and then find marginal noise regions according to the shape, length, and location of the split blocks. After the detection of marginal noise regions, removal is performed. Experimenting with a wide variety of test samples reveals the feasibility and effectiveness of our proposed approach in removing marginal noises.

Business card recognition is a difficult problem. Characters in business card are small with diverse font types. An approach using the left-right hidden Markov model is proposed for business card recognition. The hidden Markov model will output a top-10 candidate list as its recognition result. A postprocessing stage is followed to improve the recognition result. The postprocessing stage uses a bigram table as linguistic information to search for the optimized recognition result from the top-10 candidate list. Our experiments are built on the recognition of company item and address item in Chinese business cards. Bigram table and hidden Markov models are trained with a telephony database. 100 address items and 30 company items are used for testing. Experimental results reveal the validity of our proposed method.

A novel region-based multiple classifier color image retrieval system is presented. In our approach, a region-growing technique is first employed to cluster connected color pixels with the same color in an image to form color regions which are the primitive elements utilized in our proposed approach. Then, three complementary region-based classifiers that we developed are selected in the classifier selection stage, which include color classifier, shape classifier, and relational classifier. In each classifier, a virtue probability representing the probability that an image is similar to the query image is defined. Thereafter a set of virtue probabilities is calculated in each classifier. Next, the measurement dependent methods are applied to combine the virtue probabilities of classifiers in the decision combination stage. The dynamic selection scheme designed in the decision combination stage can further improve the system performance dramatically. Experimental results reveal the feasibility and validity of our proposed approach in solving the color image retrieval problem.

A discrete image flux conduction equation which is completely new in this field is proposed. The new approach starts with formulating a discrete image flux conduction equation based on the concept of heat conduction theory. Based on this discrete equation, the status change at a time point can be directly computed from its spatial neighborhood. To more accurately estimate an image flux, we have used an orthogonal wavelet basis to approximate the gradient of the intensity at each point. Since the proposed approach is discrete by nature, it is not necessary to formulate a continuous PDE to fit the discrete image data set. Furthermore, introduction of different numerical methods to solve the PDE can also be avoided. Since the proposed approach does not require that a PDE be solved, it is therefore more efficient and accurate than the conventional methods. Experimental results obtained using both synthetic signals and real images have demonstrated that the proposed model could effectively handle the selective image smoothing problem.

This paper proposes a novel run length based stroke extraction approach for Chinese characters without using the thinning method to avoid distortions around the function area generated by the method. In our approach, the run-length coding technique is first employed to obtain a special skeleton which only owns disjoint line segments with no fork points. Then, an attributed graph is constructed from the skeleton. Intersection relation among line segments is represented by a junction matrix and its associating graph. While stroke extraction is performed, the fork points can also be found. Experimental results show that the proposed approach is feasible and efficient in extracting strokes of Chinese characters.

In this paper, we present a novel region-based color image retrieval system using geometric properties. A region-growing technique is firstly employed to cluster the connected color pixels with the same color in an image to form color regions. Then, two most important descriptive geometry features are extracted. One is the spatial relational graph (SRG), another is the Fourier description coefficients (FDC) of each color region. In the matching stage, the modified relational distance graph matching between two SRG is performed firstly to find the best matches with the minimum relational distance. Then, the shape matching is applied to obtain the best vertex match with the minimum geometric distance. Experimental results reveal the feasibility of our proposed approach in solving color image retrieval problem.

The proposed system consists of two main parts: the first part is to search the potential face regions that are obtained from the triangles based on the rules of "the combination of two eyes and one mouth"; and the second part is to perform the face verification task. The proposed face detection system can locate multiple faces embedded in complicated backgrounds. Moreover, it is able to handle different sizes, different lighting conditions, varying poses and expressions, noises and defocus problems, and the problem of partial occlusion of mouth and sun glasses. Experimental results demonstrate that an approximately 98% success rate was achieved and the relative false detection rate is very low.

We present a map interpretation system for automatic extraction of high level information from the scanned images of Chinese land register maps. Our map interpretation system consists of three main components: text/graphics separation, parcel extraction, and rotated character recognition. Our approach to text/graphics separation is based on a simple yet effective rule: the feature points of characters are more compact than those of graphics. In the parcel extraction process, the proposed algorithm traces the branches between feature points to extract polygon structure from line drawings. Our character recognition method is based on the matching of extracted strokes using a neural network. The techniques of text/graphics separation and character recognition are robust to the rotation and writing style of characters. Another advantage of our separation algorithm is that it can successfully extract a character connected to a graphical line. Experimental results have shown that the proposed system is effective for the data capture of geographic information systems.

We present a novel form recognition method by analyzing the line structure embedded in an input form document. First, all vertical and horizontal lines embedded in the form image are extracted. Experimental results demonstrate the feasibility and efficiency of our proposed method in recognizing form documents.

We propose a fuzzy bipartite weighted graph model to solve fingerprint verification problem. A fingerprint image has to be preprocessed first to capture clusters of feature points. 24 attributes that are characterized by fuzzy membership functions are extracted for each feature point cluster. Input image to be verified has a set of feature point clusters that are treated as the set of left nodes in a fuzzy bipartite weighted graph, and feature point clusters of template fingerprint images are the set of right nodes in the graph. Fingerprint verification problem is then formulated as a fuzzy bipartite weighted graph matching problem that is to find an optimal matching. A matching algorithm is proposed for the fuzzy bipartite weighted graph model.

Digital video transmission across the network is now widely used in the Internet community. Visual information transmission will no doubt play an important role in a future national information infrastructure (NII). Very often, however, the channel bandwidth is fixed. Therefore, video coding based on an appropriate bit-rate control is a key to the success of accommodating visual communication in NII. In this paper, an extremely efficient rate-control algorithm for pseudoconstant bit rate video coder is proposed. The proposed technique, adaptive scene analysis, intelligently collects required information to obtain estimated bit counts from the previous, present, and future frames. This collected intelligent information is then employed to govern the resource allocation so that better objective quality can be achieved under the bit-rate restriction. Combined with a series of dynamic rate-control schemes, the proposed algorithm demonstrates impressive quality transition between successive frames. In addition, the bit-rate irregularities caused by dramatic scene transition can be smoothed out in a very short period. A comparison with the MPEG-2 Test Model 5 is also presented in this paper.

A deterministic algorithm is proposed to design the optimal stack filter. The proposed algorithm can generate the optimal stack filter in one second for a window size of 9 and it can still generate the optimal stack filter for a window size of 21 although it takes about 4 hours. Experimental results reveal the feasibility and efficiency of the proposed algorithm in finding the optimal stack filter.

Error-correcting graph isomorphism has been found useful in numerous pattern recognition applications. This paper presents a genetic-based search approach that adopts genetic algorithms as the searching criteria to solve the problem of error-correcting graph isomorphism. By applying genetic algorithms, some local search strategies are amalgamated to improve convergence speed. Besides, a selection operator is proposed to prevent premature convergence. The proposed approach has been implemented to verify its validity. Experimental results reveal the superiority of this new technique than several other well-known algorithms.

This paper presents a map interpretation system for automatic extraction of high level information (such as parcels and their attributes) from the scanned images of Chinese cadastral maps. Our map interpretation system consists of three main components: text/graphics separation, parcel extraction, and rotated character recognition. The techniques of text/graphics separation and character recognition are robust to the rotation and writing style of character. Another advantage of our separation algorithm is that it can successfully extract character connected to graphical line.

A new approach to stroke-order and stroke-number free on-line handwritten Chinese character recognition is presented in this paper. In this new scheme, the decision rule of the segment attribute is used to characterize the segment sequence appearing in each Chinese character for recognizing connected-stroke and even cursive handwritten Chinese characters. A knowledge-based radical extraction method is proposed to perform the feature extraction before radical recognition stage. The top-level and bottom-level radical classification are adopted in the coarse classification stage to reduce the number of candidate characters. In order to develop a stroke order free system, the neighboring segment matching method is proposed. Experimental results show that the proposed scheme is an efficient solution for stroke-order and stroke-number free on-line Chinese character recognition. The recognition rate is 93.4% and the recognition speed is 0.6 second per character.

A novel approach for efficient image registration is proposed. The proposed method applies wavelet transforms to extract a number of feature points as the basis for registration. From the selected feature points, a subset of possible matching pairs is selected to obtain the desired registration parameters. This subset is chosen by using the orientation difference between two target images as a criterion to eliminate spurious matching pairs. In order to predetermine the orientation difference between two target images, a so-called "angle histogram" is calculated. From the angle histogram, the orientation difference can be decided. Once the orientation difference is obtained, the desired subset can be easily determined. By randomly selecting two matching pairs from this subset, a set of registration parameters can be obtained. By checking how many matching pairs are compatible with the selected parameters, the best estimation can be determined. Compared with conventional algorithms, the proposed scheme is a great improvement in terms of efficiency as well as reliability for the image registration problem.

In this paper, a fast approach is proposed to detect and correct the skew documents. In this proposed approach, a coarse skew angle /spl theta//sub e/ is first estimated to decrease the heavily computational burden. Next, the exact skew angle /spl theta/ located in the previously estimated range [/spl theta//sub e/-3, /spl theta//sub e/+3] is then determined by making use of the Hough transform method. In addition to the skew angle detection. The skew image is quickly corrected by decomposing the traditional geometric rotation into a translation and a local rotation process. A mapping table which stores the local rotation information is employed to avoid the large floating-point operations and speed up the rotation process on general-purpose computers. Experimental results are conducted to verify the feasibility and efficiency of our proposed method.

In this paper, a concise mode is proposed to model a fundamental pattern recognition problem. From this concise mode, three optimization subproblems of pattern recognition are discussed. The reason why genetic algorithms are appropriate for solving pattern recognition problems are explained by comparing the advantages and disadvantages of various kinds of approaches. The status of applying genetic algorithms on pattern recognition is surveyed in this paper.

In this paper, we propose a recursive hierarchical scheme for radical extraction of Chinese characters. The proposed scheme, which combines structural method and statistical method, includes three modules. They are character pattern detection module, straight cut line detection module, and stroke clustering module. Experiments are conducted on 1056 constrained handwritten characters. The successful rate of radical extraction is 97.1%. The experimental results reveal the feasibility and validity of our proposed approach.

Thinning is a very important preprocessing step for many image analysis operations. The main drawbacks of traditional thinning algorithms are slow speed and serious deformation. To remedy these problems, a novel approach which generates the skeleton of binary patterns via block decomposition and contour vector matching is proposed. The performance of the proposed algorithm is compared with the existing algorithm. Experimental results confirm the superiority of the new approach.

Among various kinds of documents, forms are one of the important types. The prerequisite for form optical character recognition (Form OCR) is the extraction of characters from form documents. The authors present a clustering based technique for extracting characters from form documents. In this method, they treat the character extraction process as a pattern clustering problem. The feasibility of the novel method is demonstrated through experimenting various kinds of forms. Experimental results reveal the feasibility of the novel method.

In this paper, we propose a robust algorithm to separate Chinese characters from line drawings. This approach is based on the clustering of all feature points in the images. Using our algorithm, all Chinese characters can be completely separated from graphics without regard to the size, orientation and location of Chinese characters even if the characters touching or overlapping line problem occurs. Experiments show that our algorithm can be applied to both geographical information systems and forms processing.

A graphic searching algorithm is proposed to find the optimal stack filter. The search of the optimal stack filter is reduced to a problem of finding a minimal path from the root node to the optimal node in the error cone graph (ECG). Two graphic searching techniques, the greedy and A* algorithms, are applied to avoid the searching an extremely large search space. Experimental results are given to show the efficiency and performance of our proposed method.

This paper proposes a wavelet-based approach to solving the shape from shading (SFS) problem. The proposed method takes advantage of the nature of wavelet theory, which can be applied to efficiently and accurately represent "things", to develop a faster algorithm for reconstructing better surfaces. In order to improve the robustness of the algorithm, two new constraints are introduced into the objective function to strengthen the relation between an estimated surface and its counterpart in the original image. Thus, solving the SFS problem becomes a constrained optimization process. In the first stage of the process, the set of function variables to be solved is represented by a wavelet format. Due to this format, the set of differential operators of different orders which is involved in the whole process can be approximated with the connection coefficients of Daubechies bases. In each iteration of the optimization process an appropriate step size which will result in maximum decrease of the objective function is determined. After finding correct iterative schemes, the solution of the SFS problem will finally be decided. Compared with conventional algorithms, the proposed scheme makes great improvements on the accuracy as well as the convergence speed of the SFS problem.>

A knowledge model-based OCR system is presented for the recognition of on-line connected stroke Chinese characters. In the approach, segment attributes are first extracted to characterize the segment sequence of an unknown character. Next, radical recognition based on model matching is adopted as the coarse classification to reduce the number of candidate characters before detailed matching. Finally, a deviation modeling method is proposed to recognize not only regular writing characters but also characters with stroke-order and stroke-number deviations. The effectiveness of the approach is verified by experiments on the recognition of on-line Chinese characters.>

Searching via error code graph (ECG) to find the optimal morphological erosion filter is proposed. The problem to find the optimal solution is reduced to the problem for searching a minimal path in ECG. Since this graph satisfies some greedy properties, only a few nodes need to be traversed and examined in this graph.>

Presents a path planning algorithm to solve the find path problem in mapped environments. The moving objects considered in this paper can be any shape, which is approximated by its bounding convex polygon. Since the actual translating and rotating movement of the object in the free space is considered in the finding of a collision-free path, it won't miss those paths which require rotational maneuvering. In addition to the progressing strategy, a modified A* algorithm based on the revised depth-first search in setting up the h' value is also proposed to find the optimal path. From the experimental results, the modified A* algorithm is proved to be a very good method.>

Chinese characters are basically composed of few radicals, and it is essential to recognize the radicals inside a character before recognizing the whole character. A partial matching algorithm that is based on a modified relaxation method is proposed to do radical recognition on a scanned input character. Experiments are carried out with 204 constrained handwritten characters, each containing two to five radicals, for matching 16 radical templates. The total recognition rate is 63.3%. But the rate of identifying at least one radical in a character is 97%.>