Abstract
In this paper, a MAPSACLM algorithm for feature point matching is proposed. This method integrates the MAPSAC algorithm with nonlinear optimization by using the results of MAPSAC as the initial value of the fundamental matrix and homography matrix. Firstly, gray level cross-correlation matching method was used to realize initial matching. Secondly, the fundamental matrix and the homography matrix were estimated robustly with MAPSAC algorithm. As a result, most of the outliers were detected and removed. Then, nonlinearly optimized fundamental matrix and homography matrix by Levenberg-Marquardt algorithm were used to obtain more precise matching points. Lots of experiments show that this algorithm is efficient and it improves the robustness and accuracy of the automatic image matching in 3D reconstruction.