【摘要】 基础矩阵估计是计算机视觉的基础。提出了一种新的基于改进遗传算法的基础矩阵估计方法,它以8个匹配点作为最小匹配点集组成染色体,并利用改进8点法估计出相应的基础矩阵,同时以对极距离为核心构造适应度函数,并对传统遗传算法进行了优化改进。实验证明,此方法在估计精度和鲁棒性等方面有很大提高,易于实际应用。更多还原

【Abstract】 Fundamental Matrix Estimation is the key content of Computer Vision.A new method for estimating Fundamental Matrix based on Improved Genetic Algorithm is proposed.It uses eight points to form each chromosome as the minimum subset of correspondences,and then gets the corresponding Fundamental Matrix by improved 8-point algorithm.Besides,the fitness function is constructed based on the distance between correspondences and epipolar lines,and some genetic operators have been improved and optimized.The experiment results show that it is more precise and robust,and easier to implement.更多还原