考虑在双目立体视觉匹配算法中置信传播算法虽然能获得较好的视差图但匹配速度较慢，本文提出了一种基于置信传播的并行匹配方法。该方法以置信传播为基础，采用并行机制来提高匹配速度。首先，利用最优边缘算法计算图像中的边缘信息；然后，利用边缘信息将整个待匹配视图切割成若干小的区域，使用OpenMP多核优化算法并行对各个小区域进行匹配；最后，合并出整个视差图。使用Meddlebury测试平台提供的立体图对进行了实验。结果表明，在保证匹配准确率的前提下，置信传播算法将匹配速度提高了3.51倍，为立体匹配的实际应用奠定良好的基础。该方法在具备多核处理器的嵌入系统中也有良好的应用前景。

In this paper, a parallel stereo matching algorithm is proposed based on belief propagation algorithm to accelerate the matching speed. Firstly, the edge of the image is calculated using an optimized edge computing method, then the image to be matched is cut into several segments by the edge information, and every segment is matched based on belief propagation algorithm in parallel time. Finally, all the segment are assembled into a whole disparity image. The experimental results show that the matching speed has accelerated by 3.51 times that of traditional one in the same disparity quantity, and this method can lay the foundation for the further real-time application.