Skip to main content

Über dieses Buch

This book constitutes the referred proceedings of the 8th China Conference on Image and Graphics Technologies and Applications, IGTA 2014, held in Beijing, China, in June 2014. The 39 papers presented were carefully reviewed and selected from 110 submissions. They cover various aspects of research in image processing and graphics and related topics, including object detection, pattern recognition, object tracking, classification, image segmentation, reconstruction, etc.



Image Quality Assessment Based on SIFT and SSIM

Image quality assessment (IQA) aims to provide computational models to measure the image quality consistently with subjective assessments. The SSIM index brings IQA from pixel-based to structure-based stage. In this paper, a new similarity index based on SIFT features (SIFT-SSIM) for full reference IQA is presented. In the algorithm, proportion of matched features in extracted features of reference image and structural similarity are combined into a comprehensive quality index. Experiments on LIVE database demonstrate that SIFT-SSIM is competitive with most of state-of-the-art FR-IQA metrics, and it can achieve higher consistency with the subjective assessments in some distortion types.

Wenjun Lu, Congli Li, Yongchang Shi, Xiaoning Sun

Edge Detection in Presence of Impulse Noise

Edge detection in image processing is a difficult but meaningful problem. In this paper, we propose a variational model with



-norm as the fidelity term based on the well-known Mumford-Shah functional. To solve it, we devise fast numerical algorithms through applying the binary label-set method. Numerical experiments on gray-scale images are given. By comparing with the famous Ambrosio-Tortorelli model with



-norm as the fidelity term, we demonstrate that our model and algorithms show advantages in efficiency and accuracy for impulse noise.

Yuying Shi, Feng Guo, Xinhua Su, Jing Xu

A Novel Multi-focus Image Capture and Fusion System for Macro Photography

This paper proposes a novel multi-focus image capture and fusion system for macro photography. The system consists of three components. The first component is a novel multi-focus image capture device which can capture multiple macro images taken at different focus distances from a photographic subject, with high precision. The second component is a feature based method which can align multiple in-focus images automatically. The third component is a new multi-focus image fusion method which can combine multiple macro images to a fused image with a greater depth of field. The proposed image fusion method is based on Gaussian and Laplacian pyramids with a novel weight map selection strategy. Several data sets are captured and fused by the proposed system to verify the hardware and software design. Subjective and objective methods are also used to evaluate the proposed system. By analyzing the experimental results, it shows that this system is flexible and efficient, and the quality of the fused image is comparable to the results of other methods.

Yili Zhao, Yi Zhou, Dan Xu

Comparative Study of Near-Lossless Compression by JPEG XR, JPEG 2000, and H.264 on 4K Video Sequences

Lossless or near-lossless compression for high resolution image is required to meet the increasingly high quality demands in various application fields. With respect to this requirement, ultra-high definition image/video compression technology is researched in this papar. We first delve into the JPEG XR compression algorithm and parameters, then report a comparative study evaluating rate-distortion performance between JPEG XR, H.264 and JPEG 2000. A set of five sequences with resolution of 4K(3840x2160) have been used. The Result shows that, for the test sequences used to carry out the experiments, the JPEG XR outperforms other two coding standards in terms of the trade off between compression efficiency and hardware complexity.

Wang Dan

A Retinex-Based Local Tone Mapping Algorithm Using L 0 Smoothing Filter

In this paper, we propose a novel halo-free local tone mapping algorithm using



smoothing filter. Our method imitates the adaptation of the mechanism of the human visual system (HVS), which ensures a strong adaptability to the scenes of different dynamic ranges. Firstly, we will apply a global histogram adjustment method to the luminance image, which is a simple initial global adaptation; secondely, we will demonstrate how the luminance image is remapped by a retinex-based local tone mapping method. During the estimation of illumination, a



smoothing filter is used instead of Gaussian filter to compress the contrast while reducing the halo artifacts; and finally, through the color correction, we will show how the tone-mapped RGB image is obtained. According to our experimental results, the proposed method outperforms the state-of-art tone mapping algorithms in color rendition and detail preservation.

Lei Tan, Xiaolin Liu, Kaichuang Xue

A Fast Mode Decision Algorithm and Its Hardware Design for H.264/AVC Intra Prediction

This paper presents an architecture of mode decision algorithm for H.264/AVC intra prediction. In the algorithm design, based on the inherent correlation existing in the spatial prediction modes, a significant computational savings can be achieved. In the hardware design, through efficient sharing of configurable units and parallel executions of different candidate prediction modes, a lower hardware utilization and higher execution speed can be achieved. Synthesis results show that the proposed architecture can process HDTV (1920×1080) video at 60 fps in FPGA platform and maximum frequency achieved is 184.8 MHz.

Wei Wang, Yuting Xie, Tao Lin, Jie Hu

Research of Multi-focus Image Fusion Algorithm Based on Sparse Representation and Orthogonal Matching Pursuit

Due to the unideal effects of those common multi-source focus image fusion algorithms, in this essay we propose a multi-focus image fusion algorithm based on sparse representation and orthogonal matching pursuit (OMP), and demonstrate the results of the corresponding multi-source focus image fusion experiments by MATLAB. Compared with the fused images of the above several common algorithms by evaluating subjectively and objectively, the results suggest that the multi-focus image fusion algorithm based on sparse representation and orthogonal matching pursuit (OMP) present higher mutual information, minimum distorted values and higher Q

ab / f

values which indicate that the fused image by this algorithm can obtain more image information with a smaller distortion from the original (image?), so as to get a better image but cost much more time.

Li Xuejun, Wang Minghui

Infrared Face Recognition Based on DCT and Partial Least Squares

Infrared face imaging, being light- independent, and not vulnerable to facial skin expressions and posture, can avoid or limit the drawbacks of face recognition in visible light. However, to obtain the compact and discriminative feature extracted from infrared face image is a challenging task. In this essay, infrared face recognition method using Discrete Cosine Transform (DCT) and Partial Least Square (PLS) is proposed. Due to strong ability for data de-correlation and compact energy, DCT is studied to obtain the compact features in infrared face. To make full use of the discriminative information in DCT coefficients, the final classifier formulates PLS regression for accurate classification. The experimental results show that the proposed algorithm outperforms Principle Component Analysis (PCA) and DCT based infrared face recognition algorithms.

Zhihua Xie, Guodong Liu

Pipeline Architecture for High Speed License Plate Character Recognition

An embedded hardware for license plate character recognition is designed and implemented on an FPGA (field programmable gate array) with pipeline architecture. The architecture is based on M2DPCA (modular two-dimensional principal component analysis) algorithm. Three processing elements are contained in the proposed pipeline architecture, projection element is designed for matrix multiplication operations of feature extraction, the distances between input character and each class in training database are computed in distance element, and the nearest neighbor classification is carried out in classification element, all functions are run in pipeline. Experimental results show that very high speed is achieved, which provides approximately 28% speedup of equivalent software implementation, and also, the hardware architecture performs extremely resource economical.

Boyu Gu, Qiang Zhang, Zhenhuan Zhao

Robust Dual-Kernel Tracking Using Both Foreground and Background

The kernel-based mean shift tracker outperforms other trackers due to its innovated target representation and efficient optimization strategy. However, this representation relies overmuch on the foreground and thus, decreases the robustness to the background change and clutter. To this point, this paper presents a dual-kernel tracker based on mean shift using both foreground and background. The proposed target representation consists of foreground model and background model, and the optimizing process integrates foreground kernel iteration and background kernel iteration. Experiments indicate that the proposed tracker obtains better performance in coping with background change and clutter.

Wangsheng Yu, Zhiqiang Hou, Xiaohua Tian, Lang Zhang, Wanjun Xu

Humanoid-Eye Imaging System Model with Ability of Resolving Power Computing

This paper proposes an innovative imaging system model for human eyes with resolving power calculation, which is feasible in practice and stands on solid Physical background. The model, humanoid-eye imaging system (HIS), is constructed synthesizing an imaging component model and a photo-sensing component model based on relevant parts of human eyes. HIS integrates core features and working mechanism of human eyes and can also be regarded as simulation of various real digital imaging systems. According to criteria derived from wave optics and the theory of receptors, point resolving power for HIS is defined and its calculations are deduced as functions of specified parameters of HIS and variables of object points observed by HIS. Experiment with a camera as the application of HIS show that HIS is applicable and its resolving power calculation is precise in reality. Our work supply a novel method for the first time to efficiently connect real observing conditions with computer simulation for fields related to 3D meshes management.

Ma Huimin, Zhou Luyao

Color Cast Detection Method Based on Multi-feature Extraction

In order to raise the accuracy rate of the color cast detection and to make the method universal, the paper carries out a color cast detection method based on multi-feature extraction. Firstly, calculate the four features that are the textural property of the luminance channel, color numbers, histogram of RGB color space and statistical characteristics of the Gabor filter, then use AdaBoost to train and classify. The experiment will be done using 11346 images in the Ciurea database. The result shows that this method has a low error rate and good classification results, which is universal to natural images taken by cameras.

Minjing Miao, Yuan Yuan, Juhua Liu, Hanfei Yi

Research on an Extracting Method to Assist Manual Identification of Targets

The interpretation of the target recognition in remote sensing image is normally manually implemented by interpreters. In this essay, we propose to analyze the method for rapid extraction of suspected targets (method that can be used to assist manual interpretation) was proposed. The method is based on the image gray-scale characteristics. First of all, resolution reduction and enhancement should be used to preprocess the image. Then, through a series of means, including binarization, erosion, big target extraction and dilation, suspected targets are extracted. The experimental results show that the method can effectively and quickly extract targets from the remote sensing image which owns the features that the gray value of background area is mussy and the object area is homogeneous.

Yue Shi, Guoying Zhang

A Face Recognition under Varying Illumination

Face recognition is one of the focus studies in biometrics technology. The recognition accuracy always changes drastically in different environment, especially when it is affected by the illumination. Retinex is a method which utilizes illumination invariant, but it ignores contributions of low frequency component to face recognition. In this research paper, we propose a face recognition method based on retinex and wavelet transformmation. First, illumination invariant and variant are generated by the retinex theory. Second, decompose the illumination component via wavelet transformation and set its low-frequency coefficients to zero. In doing so, the processed illumination component is obtained by inversing the transformation. In the end, a new image is acquired by restructuring the two components. The recognition experiment will demonstrate that the proposed method ensures good performance in illumination environment.

Haodong Song, Xiaozhu Lin, Zhanlin Liu

CS-FREAK: An Improved Binary Descriptor

A large number of vision applications rely on matching key points across images, its main problem is to find a fast and robust key point descriptor and a matching strategy. This paper presents a two-step matching strategy based on voting and an improved binary descriptor CS-FREAK by adding the neighborhood intensity information of the sampling points to the FREAK descriptor. This method divides the matching task into two steps, firstly simplify the FREAK[1] 8-layer retina model to a 5-layer one and construct a binary descriptor, secondly encode the neighborhood intensity information of the center symmetry sampling points, and then create a 16-dimentional histogram according to a pre-constructed index table, which is the basis for voting strategy. This two-step matching strategy can improve learning efficiency meanwhile enhance the descriptor identification ability, and improve the matching accuracy. Experimental results show that the accuracy of the matching method is superior to SIFT and FREAK.

Jianyong Wang, Xuemei Wang, Xiaogang Yang, Aigang Zhao

Lung Fields Segmentation Algorithm in Chest Radiography

Accurate segmentation of lung fields in chest radiography is an essential part of computer-aided detection. We proposed a segmentation method by use of feature images, gray and shape cost, and modification method. The outline of lung fields in the training set was marked and aligned to create an initial outline. Then, dynamic program was employed to determine the optimal one in terms of the gray and shape cost in the six feature images. Finally, the lung outline was modified by Active Shape Model. The experimental results show that the average segmentation overlaps without and with feature images achieve 82.18% and 89.07%, respectively. After the modification of segmentation, the average overlap can reach 90.26%.

Guodong Zhang, Lin Cong, Liu Wang, Wei Guo

Automatic Classification of Human Embryo Microscope Images Based on LBP Feature

It is significant in-vitro fertilization (IVF) to automatically evaluate the implantation potential for embryos with a computer. In this essay, an automatic classification algorithm based on local binary pattern (LBP) feature and the support vector machine (SVM) algorithm is presented to classify the embryo images which will suggest whether the image is suitable for the implantation. The LBP operator is first time to be used to extract the texture feature of embryo images, and it is verified that the feature has the capacity of making two types of images linearly separable. Furthermore, a classifier based on the SVM algorithm is designed to determine the best projection direction for classify embryo images in the LBP feature space. Experiments were made with 6-fold cross validation over 185 images, and the result demonstrates that the proposed algorithm is capable of automatically classifying the embryo images with accuracy and efficiency.

Liang Xu, Xuefeng Wei, Yabo Yin, Weizhou Wang, Yun Tian, Mingquan Zhou

Robust and Accurate Calibration Point Extraction with Multi-scale Chess-Board Feature Detector

Chess-board grid has been widely used for camera calibration and the associated feature point extraction algorithm draws much attention. In this paper, a multi-scale chess-board feature point detector is proposed, along with a chess-board matching algorithm for a specific marker used in our 3D reconstruction system. Experiments show that our method is more robust and accurate compared to commonly used approaches.

Yang Liu, Yisong Chen, Guoping Wang

A Mixed-Method Approach for Rapid Modeling of the Virtual Scene Building

Virtual scene building is regarded as a significant part in cultural heritage digital presentation processes. Geometric modeling method, which has been widely used in virtual scene reconstructing, requires lots of manual operations and a long modeling period. This article proposed an approach combining three modeling methods to reconstruct 3D models. Various methods should be adopted regarding to different complexity levels of objects to make the design process more efficient and realistic. In this article, the proposed strategy will be illustrated by presenting a virtual scene of Qiao’s Grand Courtyard, one of the most famous ancient residential building groups in Shanxi province. As a result, a simple, efficient, cost-effective and high standard modeling approach has been obtained for the historic heritage digital presentation.

Pu Ren, Wenjian Wang, Mingquan Zhou, Chongbin Xu

An Improved Method of Building Rapid 3D Modeling Based on Digital Photogrammetric Technique

Building 3D modeling is a fundamental but expensive works to digital city engineer. For cutting off the workload of that, an improved rapid modeling method based on digital photogrammetric technique is proposed. Two main issues are resolved in the method. The first is how to trace and recognize the roof surfaces of a building according to its profile lines, and the second is how to create right solid models for complicated buildings usually with multiple parts. The paper gives a detailed description about the solutions and involved algorithms. Finally, multiple buildings with different roof styles are selected to test the improved method. Results show that it can get right 3D models for common buildings, and reduce the workload of delineating roof lines and possible model mistakes.

Zimin Zhang, Ying Zhou, Jian Cui, Hao Liu

3-D Reconstruction of Three Views Based on Manifold Study

Obtaining 3-D reconstruction directly and expediently for the real world has became a hot topic in many fields. A 3-D reconstruction method of three views based on manifold study is proposed. Firstly, the fundamental matrix is estimated by adjacent view and optimized under three views constraint. Then 3-D point cloud is reconstructed after getting the projection matrixes of views. Further more, benefitting from minimum spanning tree, outliers are almost excluded. To increase point cloud’s density, the optimized 3-D point cloud is interpolated based on Radial Basis Function. Afterwards, the dense point cloud is mapped to two dimensional plane using manifold study algorithm, and then divided into plane Delaunay triangle nets. Completing that, the topological relations of points are mapped back to 3-D space and 3-D reconstruction is realized. Many experiments show the method proposed in paper can achieve 3-D reconstruction for three views with quite good results.

Li Cong, Zhao Hongrui, Fu Gang, Peng Xingang

Distribution and Rendering of Real-Time Scene Based on the Virtual Reality System

There are differences between real and virtual scene information displayed on a virtual reality platform, not a true reflection of the reality of the real-time information. On the basis of virtual reality platform, how to collect, organize and publish the corresponding real-time virtual reality scene scenes information is becoming a new problem. In this essay, we will explain how to input, distribute and display real scene information based on the virtual reality system, a distributed virtual reality system. The current real-time video information corresponded with the specified scene object in the virtual reality system is recorded by the client, released to the streamer server and then distributed via content distribution server. The live streaming information displayed on the virtual reality client increases the fidelity and real-time of the virtual reality. The system is timely published, scalable and is capable of supporting remote deployment and distribution. The goal is to provide a reliable and effective real scene information dissemination and presentation of the real scene platforms based on the virtual reality environment.

Chengfang Zhang, Guoping Wang

Probabilistic Model for Virtual Garment Modeling

Designing 3D garments is difficult, especially when the user lacks professional knowledge of garment design. Inspired by the assemble modeling, we facilitate 3D garment modeling by combining parts extracted from a database containing a large collection of garment component. A key challenge in assembly-based garment modeling is the identifying the relevant components that needs to be presented to the user. In this paper, we propose a virtual garment modeling method based on probabilistic model. We learn a probabilistic graphic model that encodes the semantic relationship among garment components from garment images. During the garment design process, the Bayesian graphic model is used to demonstrate the garment components that are semantically compatible with the existing model. And we also propose a new part stitching method for garment components. Our experiments indicates that the learned Bayesian graphic model increase the relevance of presented components and the part stitching method generates good results.

Shan Zeng, Fan Zhou, Ruomei Wang, Xiaonan Luo

Dense 3D Reconstruction and Tracking of Dynamic Surface

This essay addresses the problem of dense 3D reconstruction and tracking of dynamic surface from calibrated stereo image sequences. The primary contribution of this research topic is that a novel framework of 3D reconstruction and tracking of dynamic surface is proposed, where a surface is divided into several blocks and block matching in stereo and temporal images is used instead of matching the whole surface, when all the block correspondences are obtained, a special bilinear interpolation is applied to precisely reconstruct and track the integral surface. Performance is evaluated on challenging ground-truth data generated by 3D max, and then different surface materials, such as fish surface, paper and cloth are used to test the actual effect. The research results demonstrate that this framework is an effective and robust method for dynamic surface reconstruction and tracking.

Jinlong Shi, Suqin Bai, Qiang Qian, Linbin Pang, Zhi Wang

Scene Simulation for a New Type of Armored Equipment Simulator

The scene simulation is an important content for simulator development. The fidelity of simulation directly affects the effect of the training. In this paper, first we analyse the difficulty of the scene simulation, then combined with the research of a new armored equipment simulator, we analyse and design the scene generation, signal processing and scene display. By doing this, we try to improve the close shot detail, the speed sense of movement and the sense of depth. Testified by the experiments, the training effect has been significantly improved.

Guanghui Li, Qiang Liang, Wei Shao, Xu-dong Fan

Study on the Key Technique of the Hill Shading Virtual Roaming System Based on XNA

It is helpful for people to understand the undulating topography with hill shading map. In this essay, we focus on how to shade the terrain blocks with XNA Game Studio3.1 by using the global DEM data from geographic information public service platform set as the data source. Including LOD structure, constructing a TriangleStrip terrain block model, applying HSV color model to color vertices, using XNA default lighting model and adjust relevant parameters to generate shading effect. Finally, a global height-color mapping table designed was offered, the represented effects of prototype system were also showed.

Hesong Lu, Zaijiang Tang, Qiang Liang, Wei Shao

Study on MMO in Visual Simulation Application

MMO is the network communication mechanism that has been applied in large scale network game. The related technology of MMO is quite mature, and it has relative hierarchy. However, in the DIS visual simulation filed, with the enlargement of simulation scale and the increasing of node, the load of network is much greater than before, and it will decrease the real-time of the real-time rendering at every node. In this essay, we will provide analysis of the characteristic of data exchange in visual simulation, and exolain the technical advantage of MMO on communication mechanism. Then we will present our studies on how to effectively reduce the network load, enhance the virtual scene real-time rendering the visual simulation, give a preliminary solution and conduct several proving experiment. Based on above mentioned reseach work, we built the foundation on real-time data network interaction environment that can be applied to large scale DIS visual simulation and be easily extended.

Liang Qiang, Fan Rui, Xu Renjie, Du Jun

Towards the Representation of Virtual Gymnasia Based On Multi-dimensional Information Integration

The multi-dimensional information integration of virtual gymnasia has not gained enough attention so far. It designs and implements the generation algorithms of space elements of the virtual gymnasium which efficiently solve the thorny issues, viz. the identification and the mutual map between the plan model and the stealth model of various space elements. A typical case is designed and implemented via SolidWorks, Microsoft Foundation Class (MFC), MultiGen Creator and OpenGVS based on multi-dimensional information integration where around 8000 seats in the very limited space have been designed and a seamless interaction between its plan model and stealth model is performed. Thus, users may not lose their orientations while navigating the virtual gymnasium. The research results can be applied to the on-line demonstration of sports venues, campuses and cities, etc., especially to the ticket-booking system.

Xiangzhong Xu, Jiandong Yang, Haohua Xu, Yaxin Tan

Night Vision Simulation of Drive Simulator Based on OpenGVS4.5

Night raid became more and more important in modern conflict because of the wide use of different kinds of night vision equipments; this is why the night training is playing an important role in military training. As an important method of military training, simulated training could not lose sight of night training. In this essay, we will consider the followings: firstly, the light properties of real time 3D graphics engine in OpenGVS4.5; secondly, the simulation of the vision effect of night environment, low-light-level night vision, and infrared night-vision scope; and finally, the results of our research that have been applied in a variety of driver training simulators and the satisfaying effect of the training.

Zheng Changwei, Xue Qing, Xu Wenchao

The Design and Implementation of Military Plotting System Based on Speech Recognition Technology

According to the application demand of military plotting, in this paper we design the intelligent military plotting system using speech recognition technology and MGIS, the military personnel can plot using speech. We analyze the function and performance of the mainstream speech recognition software and select the Viavoice, then we design the construction of intelligent plotting system, we mainly focus on the design of plotting and edition command, introduce the preparation work of the intelligent plotting system including data preparation and the setting of Viavoice, at last we realize the system using MGIS and Viavoice.

Wei Shao, Guanghui Li, Xiying Huang, Qiang Liang, Hesong Lu

Defect Detection in Fabrics Using Local Binary Patterns

To detect defects in fabrics more efficiently, easily and accurately, a method based on Local Binary Pattern (LBP) is proposed in this paper. The main purpose of this algorithm is to extract the feature value of fabric images. Firstly the feature of the whole defect-free fabric image is got with LBP algorithm. Then the image is divided into small detection windows, and the feature of each window can be obtained. Compare their similarity calculated by Chi-square function to get the threshold. Then process the defective images according to the same procedure. At last compare the similarity with the threshold to obtain defect regions. The defects are detected at the same time. Experimental results demonstrate that, LBP algorithm is effective in the area of detecting defects of fabrics.

Pengfei Li, Xuan Lin, Junfeng Jing, Lei Zhang

Research on Teeth Positioning Based on Wavelet Transform and Edge Detection

Human Identification from Dental X-Ray Images is a new Biometric Identification technology based on the use of modern digital image processing technology. Tooth positioning is the initial step in the individual identification system using dental X-ray image, and its main purpose is to find the accurate position of tooth in the high-resolution X-ray images, which can reduce data redundancy and provide support for the establishment of the dental images database and subsequent processing. Automated positioning and cropping of dental X-ray records is a challenging problem due to the heterogeneity of dental records. This paper proposed an algorithm to process the dental X-ray images using wavelet transform and edge detection techniques, respectively in horizontal and vertical directions to find the position of tooth area. Simulation results proved the accuracy and effectiveness of the method.

Zhou Zhou, Guoxia Sun, Tao Yang

The Orientation Angle Detection of Multiple Insulators in Aerial Image

Insulator is one of the important equipments on the transmission line, and its orientation angle detection is an important preprocessing step for accurate localization of insulator. This paper proposes a method of orientation angle detection of insulators in order to realize the orientation determination of multiple insulators in aerial image with complicated background. Firstly, extract sequential edge points and define their orientation angles of each linking contour in the preprocessed image. And secondly define the points with sign-Changing angle as candidate points. Finally, the accurate points can be picked up by RANSAC (RANdom SAmple Consensus) from candidate points, and the straight line which accurate points locate in is the main direction of the target. Experimental results verify that the proposed method has higher detection accuracy compared with the existing methods that can lay the foundations for the localization of multiple insulators in complicated background.

Zhenbing Zhao, Ning Liu, Mingxiao Xi, Yajing Yan

Improved Robust Watermarking Based on Rational Dither Modulation

Rational dither modulation (RDM) watermarking was presented to resist amplitude scaling attack. This property is achieved by quantizing the ratio of consecutive samples instead of samples themselves. In this paper, we improve the performance of basic RDM watermarking to resist more types of watermarking attacks. We improve the robustness of our modified RDM watermarking by the following three aspects: 1) The quantization step size is increased by modifying two coefficients instead of only one coefficient in the basic RDM method, 2) Several modification rules are defined to reduce embedding distortion, and 3) The coefficients with larger magnitudes in the lowest sub-band in DWT domain are selected to embed watermark. A variety of attacks are implemented to evaluate the performance of our method. Experimental results demonstrate that our method outperforms the basic RDM method and two state-of-the-art watermarking methods over a wide range of attacks and it also has good imperceptibility.

Zairan Wang, Jing Dong, Wei Wang, Tieniu Tan

The Research of Vehicle Tracking Based on Difference Screening and Coordinate Mapping

In order to identify vehicle driving cycle by monocular camera and then offer automotive active safety systems such as ACC (Adaptive Cruise Control) system, accurate condition identification signal, a difference screening method based on haar feature is put forward to identify the vehicle and a method based on coordinate mapping is improved to eliminate the impact that the changes of pitch angle make on the accuracy of positioning the vehicle, then combine with Karman filter technology to track vehicles. Finally, the studied method is used to track the vehicle in an actual video and the test results show that the method can correctly identify the image of vehicle, and accurately track the spatial position of vehicle. As a result, the studied method can be used to offer an active safety system like ACC accurate condition identification signal.

Zhang Jun-yuan, Liu Wei-guo, Tong Bao-feng, Wang Nan

Pathology Image Retrieval by Block LBP Based pLSA Model with Low-Rank and Sparse Matrix Decomposition

Content-based image retrieval (CBIR) is widely used in Computer Aided Diagnosis (CAD) systems which can aid pathologist to make reasonable decision by querying the slides with diagnostic information from the digital pathology slide database. In this paper, we propose a novel pathology image retrieval method for breast cancer. It firstly applies block Local Binary Pattern (LBP) features to describe the spatial texture property of pathology image, and then use them to construct the probabilistic latent semantic analysis (pLSA) model which generally takes advantage of visual words to mine the topic-level representation of image and thus reveals the high-level semantics. Different from conventional pLSA model, we employ low-rank and sparse matrix composition for describing the correlated and specific characteristics of visual words. Therefore, the more discriminative topic-level representation corresponding to each pathology image can be obtained. Experimental results on the digital pathology image database for breast cancer demonstrate the feasibility and effectiveness of our method.

Yushan Zheng, Zhiguo Jiang, Jun Shi, Yibing Ma

Remote Sensing Image Change Detection Based on Low-Rank Representation

In this paper we propose an unsupervised approach based on low-rank representation (LRR) for change detection in remote sensing images. Given a pair of remote sensing images obtained from the same area but in different time, the subtraction and logarithm ratio operators are firstly applied to obtain two difference images. Meanwhile the sparse part generated by LRR is also employed for acquiring another difference image, which can detect the change information. Afterwards, LRR is used again to obtain the low-rank part of these three difference images which can reflect the common characteristics. Finally


-means is performed on the low-rank part and thus the final result of change detection can be gained. Experimental results show the effectiveness and feasibility of the proposed method.

Yan Cheng, Zhiguo Jiang, Jun Shi, Haopeng Zhang, Gang Meng

Two New Methods for Locating Outer Edge of Iris

The two new methods for locating the outer edge of iris were presented by researching of iris images. The first is to locate the inner edge of iris by using canny operator and Hough transform method. The selection of the appropriate threshold is based on the diagram which observes the color distribution of the iris image; the second is that the circle integral and linear segment methods to locate the outer edge of iris are presented. We would like to indicate that the experiment results show that both of the two methods have the following qualities: rapidity, availability, real time calculations and accuracy in locating the iris outer edge.

Yujie Liu, Hong Tian

Effects of the Gridding to Numerical Simulation of the Armour-Piercing Warhead Penetration through the Steel Target

paper studies the influencing rule of gridding definition forcalculating resultin numerical simulation of the armour-piercing warhead penetration through the steel target, so we can get the reasonable scope of grid size. Paper has used the adopted explicit dynamic analyzing program AUTODYN to simulate the process of the armour-piercing penetration through the steel target. Based on the simulation, we get the penetration deepness,penetration overload and damage area of the target in conditions of different gridsize. In order to get more reasonableresult, the scope ofgridsizeis equaled to about 5.0mm by contradistinguishing and analyzing.

Jun-qing Huang, Ya-long Fan Rui Ma, Wei Shao


Weitere Informationen