We propose a novel post-processing approach for the local optimization of Locally Optimized RANdom SAmple Consensus (LO-RANSAC), called the Multi-Estimation-based Parameter Centroid (MEPC) decision. It is observed that the optimal thresholds for hypothesis generation and evaluation differ in local optimization with the inner RANSAC. Instead of binary labeling for inliers and outliers, a new ternary labeling for inliers, midliers, and outliers is introduced, using two thresholds. Our experimental results show that the highest-scoring model measured by the ternary method is closer to the real model than that measured by the existing binary method. However, it should be noted that the highest score still does not correspond to the best model due to inaccurate evaluation by data noise. We introduce a new linear model centroid decision method to compensate for the highest-scoring model distorted by noise. In this process, an efficient method for measuring the similarity between two hypotheses is introduced, and candidates close to the real model are found by comparing their similarity with the highest-scoring model. Our approach determines a representative model of the multiple candidate hypotheses, which is defined as the geometric centroid of hyperplanes. We test on various datasets for homography, fundamental, and essential matrices, demonstrating that applying MEPC to existing RANSAC algorithms achieves more accurate and stable model estimation. Moreover, additional experiments on vanishing point detection show the potential of our approach for various model estimation applications.

Most computation-intensive industry applications and servers encounter service-reliability challenges due to the limited resource capability of the edge. Achieving quality data fusion and accurate service reliability with optimized service-x execution cost is challenging. While existing systems have taken into account factors such as device service execution, residual resource ratio, and channel condition; the service execution time, cost, and utility ratios of requested services from devices and servers also have a significant impact on service execution cost. To enhance service quality and reliability, we design a 2-step adaptive service-X cost consolidation (ASXC2) approach. This approach is based on the node-centric Lyapunov method and distributed Markov mechanism, aiming to optimize the service execution error rate during offloading. The node-centric Lyapunov method incorporates cost and utility functions and node-centric features to estimate the service cost before offloading. Additionally, the Markov mechanism-inspired service latency prediction model design assists in mitigating the ratio of offload-service execution errors by establishing a mobility-correlation matrix between devices and servers. In addition, the non-linear programming multi-tenancy heuristic method design help to predict the service preferences for improving the resource utilisation ratio. The simulations show the effectiveness of our approach. The model performance is enhanced with 0.13% service offloading efficiency, 0.82% rate of service completion when transmitting data size is 400 kb, and 0.058% average service offloading efficiency with 40 CPU Megacycles when the vehicle moves 60 Km/h speed around the server communication range. Our model simulations indicate that our approach is highly effective and suitable for lightweight, complex environments.

Computation and memory resource management strategies are the backbone of continuous object tracking in intelligent vehicle orchestration. Multi-object tracking generates enormous measurements of targets and extended object positions using light detection and ranging (Lidar) sensors. Designing an adequate object-tracking system is a global challenge because of dynamic object detection and data association uncertainties during scene understanding. In this regard, we develop an intelligent multi-objective tracking (IMOT) system with a novel measurement model, called the box data association inflate (BDAI) model, to assess each target’s object state and trajectory without noise by using the Bayesian approach. The box object filter method filters ambiguous detection responses during data association. The theoretical proof of the box object filter is derived based on binomial expansion. Prognosticating a lower-dimension object than the original point object reduces the computational complexity of vehicle orchestration. Two datasets (NuScenes dataset and our lab dataset) are considered during the simulations, and our approach measures the kinematic states adequately with reduced computation complexity compared to state-of-the-art methods. The simulation outcomes show that our proposed method is effective and works well to detect and track objects. The NuScenes dataset contains 28130 samples for training, 6019 examples for validation and 6008 samples for testing. IMOT achieves 58.09% tracking accuracy and 71% mAP with 5 ms pre-processing time. The Jetson Xavier NX consumes 49.63% GPU and 9.37% average power and exhibits 25.32 ms latency compared to other approaches. Our system trains a single pair frame in 169.71 ms with affinity estimation time of 12.19 ms, track association time of 0.19 ms and mATE of 0.245 compared to state-of-the-art approaches

Roadside Unit (RSU) plays a significant role in Vehicle-to-Everything (V2X) service to accomplish adequate performance for the autonomous driving system. Haphazard RSU deployment and selection cause to diminish the frame-work efficiency and fail to meet the service capability of most vehicles. Reducing latency by timely delivering service is a phenomenal concern of edge orchestration to enhance the quality of service (QoS) with high reliability. With this motivation, this study focuses on maintaining a trade-off between RSU deployment and selection to achieve a high convergence rate of service offloading policy for effective service consolidation. An RSU-Assisted Service Consolidation (RASC) approach is proposed considering probabilities of three service demand rates, Computation-intensive service rates, and resource prediction rates based on Deep Reinforcement Learning. A resource-weight factor is designed to classify and map the arrived services to the potential RSU based on a discrete concave-convex procedure to avoid the execution hiccups for achieving a high convergence rate of service offloading policy. The average probability rate helps to select the potential RSU to fulfill the requirements of delay-sensitive service requests. The simulation results shows the RASC method improves the service reliability rate by at least 69.5%, the energy preservation rate by 27.1%, and the service offloading reduction rate by 73.08% compared to state-of-art schemes.

Role of RSUs become essential in enhancing the service reliability rate of the end vehicle to strengthen autonomous vehicle technology. Computation-intensive services are delay-sensitive, and most existingmethodsattempted comprehensively to meet the application deadline but have not reached the expectations due to classical computing. In this regard, we design a novel offloading decision-making method based on quantum theory through reinforcement learning. Grover's algorithm is employed to select a feasible device based on cost and energy usage probability ratio. Theoretical and mathematical validations and simulation outcomes confine the impact of novel decision-making methods on the statistical constraints of the heterogeneous framework.

Cyber-physical systems (CPS) consist of a variety of multicore architectures, including central processing units (CPU) and graphical processing units (GPU). In general, programmers assign sequential programs to the CPU while parallel applications are assigned to the GPU. This article provides a method for mapping an OpenCL application to a heterogeneous multicore architecture using active fuzzy learning to determine the adequacy and processing capabilities of the application. During learning, subsamples are created by developing a machine learning-based device suitability classifier that predicts which processors would have excessive computational compatibility for running OpenCL programs. In addition, this study integrates an active learning model based on entropy with a fuzzification model to find nonoverlapping patterns. To minimize rule generation, the fuzzification-based weighted probabilistic technique is presented. The defuzzification process is optimized by using uncertainty values in conjunction with classification probability. In addition, 20 different features are proposed for extraction using the newly developed LLVM-based static analyzer. The correlation analysis approach is used to determine the optimal subset of features. The synthetic minority oversampling approach with and without feature selection is used to differentiate the class imbalance problem. Instead of manually modifying the machine learning classifier, a tree-based pipeline construction approach is used to determine the optimal classifier and associated hyperparameters. Experiments are then conducted on a set of benchmarks to verify the performance of the designed model. The results show that by increasing the number of training examples and including an entropy uncertainty measure, the proposed model is able to support and improve decision boundaries. We achieved a high F-measure of 0.77 and a ROC of 0.92 by optimizing and reducing the feature subsets.

In this paper, we propose a constrained H∞ controller design method for the active suspension systems (ASS) with aperiodic sampling. The ASS are controlled by a hydraulic actuator which is usually limited by physical constraints. To utilize the information of saturation level, we express the ASS as the linear parameter varying (LPV) system by adopting time-varying parameters which depend on the constrained input. For the implementation of digital controllers, we consider the sampled-data H∞control. The stabilization condition is derived by using looped functional approach. The proposed controller guarantees the desired performance of the ASS despite of aperiodic sampling and input saturation. To demonstrate the validity of the proposed method, the simulation results are compared with the case of passive suspension systems.

Continuous-time semi-Markovian jump neural networks (semi-MJNNs) are those MJNNs whose transition rates are not constant but depend on the random sojourn time. Addressing stochastic synchronization of semi-MJNNs with time-varying delay, an improved stochastic stability criterion is derived in this paper to guarantee stochastic synchronization of the response systems with the drive systems. This is achieved through constructing a semi-Markovian Lyapunov-Krasovskii functional together as well as making use of a novel integral inequality and the characteristics of cumulative distribution functions. Then, with a linearization procedure, controller synthesis is carried out for stochastic synchronization of the drive-response systems. The desired state-feedback controller gains can be determined by solving a linear matrix inequality-based optimization problem. Simulation studies are carried out to demonstrate the effectiveness and less conservatism of the presented approach.

A typical road environment includes a variety of moving/stopping objects. It is important to detect moving objects for safety driving. However, moving object detection is difficult, as background is also moving when ego-vehicle is driving. In this paper, we introduce the method that detects moving objects using fundamental matrix. Fundamental matrix is calculated from previous and current frames. Feature points are obtained using Harris corner detection in previous frame and correspondence points in current frame are estimated by Lucas-Kanade method. The feature points are evenly selected by considering x-coordinate, y-coordinate and feature magnitude of flow for calculating fundamental matrix using RANSAC. Next, whether each point is inlier or outlier is determined and the points with negative direction is detected. In the simulations, we compare the performances according to the selecting method.

Images projected from HUD (Head-up Display) are always overlaid with moving background scene. In such cases, drivers suffer from understanding instinctively the HUD information. Visibility enhancement of HUD information is very important for drivers to recognize easily and quickly the information. In this paper, we propose two visibility enhancement methods of HUD image, which are based on the dominant color of background scene. One is to change the color of symbols, another is to add different color outline to symbols of HUD information, depending on the similarity between HUD information and background dominant colors. To compare the color similarity, HSI (Hue, Saturation, Intensity) color domain is used. The proposed methods are designed to minimize the color change, so as to maintain the meaning of symbol color. Through the simulations, we show that the proposed color changing and outline methods are effective to improve the instinctive understanding.

In this paper, we propose a full reference (FR) video quality assessment (VQA) algorithm for the broadcasting multimedia signal. The relationships between the VQA parameters and the quality score obtained by the subjective human visual system (HVS) model of the distorted videos are investigated. Considering the high correlations and the importance of the VQA parameters, a new FR VQA algorithmis worked out to estimate the quality of the video sequences. The simulation results show that our proposed algorithm can represent the quality of the video signals which broadcasted through network efficiently and give high correlation with the subjective video quality (VQ) score.

We propose a new method for watermarking of 3D surface mesh. After irregular wavelet analysis, the watermark is embedded in the scale coefficients. We modify the mean of the distribution of the vertex norm of the approximation mesh in order to obtain blind scheme which does not require the original mesh for detection. Experimental results show the effectiveness of the proposed algorithm both in terms of invisibility and robustness against simplification and progressive lossy compression

In this paper, we present a wavelet-based progressive compression method for 3-D dynamic meshes. Our method exploits the spatial and temporal redundancy. We encode the geometry of base mesh, the wavelet coefficients and the connectivity of each resolution level in order to reduce the spatial redundancy of intra meshes. For inter mesh coding, we encode the differences of geometry of base meshes and of their wavelet coefficients between adjacent frames to reduce the temporal redundancy. Our proposal is based on the wavelet-based multiresolution analysis which uses a perfect reconstruction filter bank and therefore it enables not only progressive representation but also lossless compression. The simulation results demonstrate that the proposed method is applicable to lossy and lossless compression of 3-D dynamic meshes.

In this paper we revisit the triangulated cubes of the marching cubes algorithm in order to regularize the connectivity of the iso-surface mesh i. e. to maximize the valence six vertices. This leads to both an enhancement of the mesh topology and a significant reduction of the connectivity coding cost. We prove the effectiveness of the approach by processing medical examples with various geometry and topology.

We propose a discrete wavelet transform (DWT) based watermarking method, which can be conveniently integrated in the forthcoming JPEG-2000 baseline. Conventional DWT-based watermarking techniques insert a watermark into the coefficients after the transform is completed, while the proposed method inserts a watermark into the coefficients obtained from the ongoing process of lifting for DWT. The proposed method allows us to determine selectively frequency characteristics of the coefficients where the watermark is embedded, so that the watermark cannot be easily removed or altered even when filter-banks for DWT are known. Through simulations, we show that the proposed method is more secure and more robust against various attacks than the conventional DWT-based watermarking.

UNLABELLED: PhyloDraw is a unified viewing tool for phylogenetic trees. PhyloDraw supports various kinds of multi-alignment formats (Dialign2, Clustal-W, Phylip format, NEXUS, MEGA, and pairwise distance matrix) and visualizes various kinds of tree diagrams, e.g. rectangular cladogram, slanted cladogram, phylogram, unrooted tree, and radial tree. By using several control parameters, users can easily and interactively manipulate the shape of phylogenetic trees. This program can export the final tree layout to BMP (bitmap image format) and PostScript. AVAILABILITY: http://pearl.cs.pusan.ac.kr/phylodraw/ CONTACT: [email protected]

We propose a wavelet based local subdivision of 3-D surfaces, which can be effectively applied to a mesh with complex and high curvature faces. The proposed approach is an extension of M. Lounsbery's works (1994), which has been developed only for regular triangular mesh subdivision. For such a purpose, a bi-orthogonal wavelet basis is constructed by defining a local inner product and using the lifting scheme. As only one filter bank is used for the global and local analysis of the surface and also one filter bank for synthesis, this method is very effective. Through the computer simulation tested on some example mesh, we show that the proposed local subdivision is very effective.

We propose a new subdivision scheme derived from the regular 1:4 face split, allowing analysis of irregularly subdivided triangular meshes by the wavelet transforms. Some experimental results on real medical meshes prove the efficiency of this approach in multiresolution schemes. In addition we show the effectiveness of the proposed algorithm for lossless compression.

In this paper we introduce a new compression technique for the connectivity of a Triangulated Irregular Network (TIN) by using Delaunay triangulation. The strategy in this paper is based on that most of 2.5 dimensional terrain connection are very similar to plane-projected Delaunay triangulation of their vertex set. We only need to keep the vertex coordinates and a few edges that are not included in the Delaunay edges. We present an efficient encoding method for the set of edges using vertex reordering and general bracketing algorithm. Also we propose progressive visualization method for web-based GIS. In the experiments, the method was examined on several TIN data with various resolutions, which were generated by five typical terrain simplification algorithms. The connectivity information of common terrain data could be compressed by 0.23 bits per vertex.

A new transform, called the rounding transform (RT), is introduced in this paper. This transform maps an integer vector onto another integer vector by using weighted average and difference filters followed by a rounding operation. The RT can be applied to lossless pyramid structured coding with various elementary block sizes and filters. In addition, it generalizes other mean based lossless pyramid structured coding schemes.

In this paper, a new approach is proposed, the rounding transform (RT) based lossless subband coding. For the such purpose, an extended version of RT, namely overlapping RT (ORT), is defined in the Z-transform domain and used to develop the lossless subband coding system. The main idea is that the polyphase matrix of the analysis filter bank is decomposed into several ORT matrices of size 2 by 2. Compared to the previous scheme (the lifting), the proposed scheme gives not only simpler mathematical representation, but also allows various possibilities when a given filter bank is implemented. As examples, five filter banks are applied to lossless image compression by the proposed ORT and their performances are compared in terms of the total lossless first order entropy.