Liu Binghan

Moving shadow detection based on Susan algorithm

In Intelligent Transportation Systems, moving shadows have always been wrongly detected as foreground objects, thus causing bad effect on the latter targets tracking and identify. In order to subtract the shadows lie in different circumstances,there has put forward a method of moving shadow detection based on Image edge detection which using Susan algorithm[9]. After getting the background with mixture Gaussian distribution, the moving foreground position can be accurately obtained via the background subtraction method, then use the Susan algorithm to detect the image edge in interested area. Finally, do shodow detection by analyzing the statistic characteristics of edge pixels. According to the experiment, this method is easy to operate and possesses high rate of accuracy, low rate of complexity, and well adapt to different kinds of shoadow distribution.

Automatic Recognition Algorithm of Traffic Signs in Road Tunnel

We conduct feature extraction and feature selection of the pattern of traffic signs based on environmental characteristics of the road tunnel, and the color and shape information of traffic signs, then further accomplish multi-level classification of traffic signs using decision tree method. The method proposed in this paper based on decision tree classification algorithm can convert a complex multi-class problem into several simple classifications. Experimental results show that the algorithm has good recognition results.

A method of feature selection based on Particle Swarm Optimization algorithm with trans-gene operator

The purpose is to apply the binary particle swarm optimization (BPSO) in feature selection. According to feature selection, the method of particles coding, fitness functions, and feature selection functions were designed. Furthermore, trans-gene operator was adopted to solve BPSOpsilas premature convergence. The simulation experiment results show that the feature subsets selected by this new algorithm are representative. The conclusion is that this algorithm is available and feasible in feature selection.