This paper makes a concise introduction to seven classic algorithms of triangular mesh subdivision, and makes a classification and comparison between them according to their continuity, own advantage and application status. In order to improve the visualization of triangular mesh subdivision, interactive display control is implemented by using MFC and OpenGL while state machine model based on functional class is used as the software operating pattern. On the basis of this, the prototype implementation of Loop algorithm is performed. The improved method of solving existing problematic issues from prototype implementation is presented.

In order to track a target in space and scale in a complex background,a target tracking algorithm based on the nonparametric clustering and multi-scale images is presented.In this algorithm,first,a modified nonparame-tric color-clustering method is employed to adaptively partition the color space of a tracked object,and the Gaussian function is used to model the spatial information of each bin of the color histogram.Next,the Bhattacharyya coefficient is adopted to derive a function describing the similarity between the target model and the target candidate.Then,a coarse-to-fine approach of multi-scale images is employed to implement the spatial location of the tracked object.Finally,the derived automatic bandwidth selection method of kernel function is applied to obtain the scale of the tracked object.Experimental results show that the proposed algorithm outperforms the classical mean shift tracker.

The former symmetric encryption algorithms use substitution and transposition techniques.This paper presents a symmetric encryption algorithm based on circle property.The techniques of random number and cryptographic hash function are used in this algorithm.So the cipher text randomly distributes in the n-dimensional space and led to cryptanalysis being difficult.From the security analysis the brute-force attacks are infeasible.It is applicable for encryptions with timestamp and short plain text encryptions.Experimental results verify that this algorithm is feasible.

For most video text is rich in edge,uniform-colored and horizontally ranged,the candidate text areas in video image are determined by using a fast deriche edge based algorithm,and the accurate binary text images are extracted from the areas by using a color based algorithm.The experimental results show that the method is effective in text extraction in complex-background video frame,and has higher processing speed and better extraction result in comparison with color alone based method.

A blackboard-based system for human detection in an indoor environment is designed and implemented,in order to accomplish human detection in wide angle camera images. The algorithm of this system excludes other nonhuman objects to detect human. Experimental results show that the blackboard-based method for human detection is feasible. It is applicable even in the presence of partial occlusion of humans,rotation of the head and variation of skin colors. It gives a feasible method for human detection in wide angle camera in indoor environments.

An automatic method of human body matting against a constant colored background was presented.First,edge detecting and the background features were used to estimate the imprecise region of the body,and then the region was divided into different parts using the body shape features.Based on the information of the different regions,the body edge was matted by using the distance proportion of the color between the current pixel and background.In addition,the method could effectively mate body edge from the images where the color of body's clothes was similar to that of the background.

As most existing document retrieval models are inefficient in semantic learning and are unable to learn the document similarity in topic level, a topic-based document retrieval model (TDRM) is proposed in this paper. TDRM provides a common topic space for all documents, represents each document as a vector in the common space, defines the document similarity as the cosine of the angle between document vectors, and uses Latent Dirichlet Allocation to learn the topic distribution of each document. Experimental results show that, as compared with the document similarity model based on the TextTiling and the optimal matching of bipartite graph, TDRM is of higher average precision and recall in the retrieval of similar document, with its harmonic mean of average precision and recall being 44% greater than that of the reference model.