一种基于极线约束的最小二乘匹配方法研究

Least Squares Image Matching Method Based on Epipolar Constraint

图像匹配技术在图像处理与计算机视觉中有着非常重要的作用。最小二乘影像匹配方法是常用的匹配方法之一。本文提出了一种基于极线约束的最小二乘图像匹配算法。该算法将极线约束和最小二乘法结合实现图像匹配。首先利用两幅图像间的极线约束得到多个预匹配点,确定预匹配区域,然后用最小二乘法求得最佳匹配。实验结果表明,该方法效果良好,具有实用价值。

Image matching technology has a very important role in image processing and computer vision. The Least squares image matching method is one of the commonly used ways. This paper proposes the least squares image matching algorithm based on epipolar restriction, which uses epipolar restriction and least squares comprehensively. Firstly, find some reliable points with epipelar constraint, determine initially matching region. Then use the least square method to achieve the best matching. Experi-mental results show that, this algorithm is effective and practical.