复杂背景下立体图像的预处理和数据净化

面向具有复杂背景的立体图像分析，针对其中的图像预处理和数据净化问题进行研究，提出了一套用于提取图像中目标区域的算法。实验结果证明，该算法能快速、有效地消除立体图像中的无用背景信息，降低图像数据的冗余，对提高后续分析的速度及结果精度有帮助。

Preprocessing and Data Segmentation of Binocular Images with Complicated Background

This paper targets the problems of preprocessing and data segmentation of stereo images with complicated background. It proposes a group of algorithms to separate obstacle areas from complicated changing background. Experiments show that the algorithms are fast and efficient in reducing the redundancy of the original data, and thus improve the speed and accuracy of the consequent stereo vision analyses.