Applied Mechanics and Materials Vols. 427-429

Abstract: multiple object tracking is an active and important research topic. It faces many challenging problems. Object extraction and data association are two most key steps in multiple object tracking. To improve tracking performance, this paper proposed a tracking method which combines Kalman filter and energy minimization-based data association. Moving objects are segmented through frame difference. Its can be consider as the vertex. All detections in adjacent frames are be used to construct a graph. The energy is finally minimized with a graph cuts optimization. Data association can be consider as multiple labeling problems. Object corresponding can be obtained through energy minimization. Experiment results demonstrate this method can be accurately tracking two moving objects.

Abstract: Direction of Arrival (DOA) estimation is of big importance in Sonar, Radar and satellite systems. In this work, we estimated the 2-D DOA (two-dimensional DOA) of coherent wide-band sources by using a broadband focusing method, which was based on the separation of angle and frequency of the array manifold matrix. At first, we transform the element position of the plane array into the spherical coordinates. Then we adopted a generalized Fourier series expansion based on the spherical harmonics and the first kind of spherical Bessel functions to get the arbitrary plane array steering vector. In this vector, the angle and frequency of array manifold matrix was separated, so we could get the matrix of focusing transformation. Finally, we used the traditional narrow-band algorithm to estimate the 2-D DOA. The design is suitable for arbitrary array geometry models of coherent broadband signals, and avoids the angle pre-estimation in traditional methods. Simulation results demonstrate its validity and better performance.

Abstract: This paper proposes an extending dynamic range method for infrared finger vein image acquisition by multi-light-intensity imaging. The captured images contain high dynamic range by different light intensity irradiating. The fast switch of light-intensity is synchronous with the camera imaging and the current image illumination analysis, these ensure the light intensity is automatic regulated and the captured images are in pixel-level registration. The experiment shows the proposed acquisition method is effective and the captured multi-light-intensity finger vein images are convenient for image fusion and subsequent processing.

Abstract: Iamge repair using the digital image processing technology has become a new research point in computer application. A novel method of local statistic enhancement based on genetic algorithm is proposed in this paper for the image enhancement. The modified amplified function are used as the jugement criterion, and the optimal paremeters are searched by the genetic algorithm. Experimental results show that the quality of images is improved dramatically by using this method.

Abstract: By studying the Chinese number and quantifier prefix (CNQP) as a special language phenomenon in machine translation, this paper presents a CNQP recognition method, which is rule based and independent of word segmentation. The method expressed CNQPs compositions using Backus-Naur Form (BNF), and took the numeral as the active information and the quantifiers as the boundaries of the CNQPs. To avoid the word segmentation noise, a forward maximum matching method was used for obtaining the compositions of the CNQPs, which can be fed into the statistical parser for the analysis of the Chinese sentences. The experimental results indicate the proposed method as a pre-processing module can effectively improve the parsing results of the statistical parser without retraining on experimental data constructed manually, which can further enhance the translation qualities.

Abstract: In this work, we develop a simple yet practical algorithm for solving nonlinear optimization problems by finding a root of a real function with a good local convergence. The algorithm can be easily implemented in software packages for achieving desired convergence orders. For the general-point formula,the order of convergence rate of the presented algorithm is , the unique positive root of the equation .

Abstract: Block Compressive Sensing (BCS) is a image reconstruction model based on CS theory. By use the same measurement matrix to obtain the data in the form of Block × Block. Algorithm meaning to solve the problem that the traditional CS measurement matrix required for large storage, but different block has important influence on reconstruction time and effect. In this paper, find out the optimum parameters of the block. By compared the PSNR and reconstructed image effect under different sampling rate and different block sizes.

Abstract: The edges are the most fundamental and important characteristic of an image, edge detection is the key link and classic topic in machine vision and image processing. Considering the human visual characteristics, the intrinsic gradient direction information of color images was used to obtain the pseudo-color edges of the images by conducting multichannel edge detection. After enhancing the edge information and removing the correlation, the brightness is extracted in order to obtain complete image edge information. To make the edges more smooth and continuous, the Hessian matrix is used to remove coarse edges and edges with redundant background texture. The experiment verifies the effectiveness of the proposed algorithm, and the comparison with other scheme indicates that our scheme can improve the effectiveness, continuity and sharpness of edge detection.

Abstract: This paper describes a post-processing method for Chinese and Japanese character recognition based on dictionary. By the analysis results of recognition in the processing of OCR, we can find some segmentation and recognition errors do not conform to the rules of lexical and just recognized as the characters which its fonts approach to the scanned texts. For these errors we can deal with them by the Fix Length Segmentation Matching based on Dictionary and the Glyph Code Matrix Transforming. Through the above processing, most of the inaccurate recognitions can be corrected and by the experimental results, it can be proved that this method is an effective way to improve the recognition rate of Chinese and Japanese Character.

Abstract: The image of the underwater target was gained by simulating the echo sound of the target. And the target strength was given in the image varying with the direction and distance from the receiving array. The length of the target can be shown in the image. So with information of the target strength and length, the characteristic of the target can be got and the detection of the target can be done.