会议专题

Image Processing and Modeling of Virtual Scenes based on nonlinear ICA

The intdligent role in virtual environment can be located when using vision to get the target image, but there is a source separation problem in case of unknown complex natural signals. In order to resolve the source separation problem, this paper proposed the target threedimensional modeling based on the visual image processing of virtual robot; applied the method based on the nonlinear Principle component analysis indicator and style of learning rules of ICA to separate the signal image, and put forward the mathematical model and rule, then introduced the feature extraction of the signal image and took the natural Litchi as an example to descript the ffectiveness. Finally, the applications of ICA for Virtual Scene Modeling were indicated.

Principle component analysis Style of learning rules Signal separation Feature extraction Visual image processing

Chen Yan Zou Xiangjun Chen Haolei Chen Lijuan

Key Laboratory of Key Technology on Agricultural Machine and Equipment, College of Engineering, South China Agricultural University, Guang Zhou,510642, China

国际会议

2010 International Conference on Measuring Technology and Mechatronics Automation(ICMTMA 2010)(2010年检测技术与机电自动化国际会议)

长沙

英文

1543-1546

2010-03-13(万方平台首次上网日期,不代表论文的发表时间)