基于蚁群优化算法的立体匹配

立体匹配技术使得通过像点获取景物的距离信息，实现三维立体再现成为可能，是计
算机视觉研究中最基本的关键问题之一。本文选择图像的边缘点作为匹配基元，以边缘特征点处的灰度值、梯度的大小和方向、拉普拉斯值作为其属性值，依据立体匹配的约束条件，建立能量函数。在进行图像的立体匹配的过程中，运用蚁群优化算法找寻使能量函数达到最小的路径，从而实现立体匹配。实验证明，该方法具有较强的稳定性，能得到较高精度的匹
配结果。
关键词：立体匹配；能量函数；蚁群优化算法

Abstract: Stereo matching makes it possible to reproduce a three- dimensional stereo by getting the distance information through pixels. It is one of the fundamental and significant problems in the study of the computer vision. In this paper, choosing the edges as the primitives for matching, the grey values of the edges and the magnitude and direction of the edge gradient and the Laplace values were figured out as the properties of the edge feature
points, according to the constraints for stereo matching, the energy function was built.
During the process of stereo matching, ant colony optimization algorithm was applied
to this problem for finding the route minimizing the energy function. Experiment
results show that this algorithm is more stable and it can get the matching result with
better accuracy.
Key words: Stereo matching; Energy function; Ant colony optimization algorithm