针对光照分布不均匀的室内环境下低动态载体速度计算实时性较差的问题,提出一种基于改进奇异值分解(SVD)-Harris的低动态载体速度快速计算的新方法。利用SVD对相邻两帧视觉图像分别进行压缩与重构,并结合改进的Harris角点检测算法对两帧图像进行特征点的检测;利用归一化互相关(NCC)模板匹配算法对相邻两帧视觉图像的特征点进行粗匹配;利用随机抽样一致性算法进行误匹配点对的剔除;利用特征匹配点对的信息对载体的速度进行计算。实验结果表明:传统算法的平均计算时间为3.07 s,而改进算法的平均计算时间为0.71 s,且传统算法的误匹配率远大于改进算法。与传统的NCC模板匹配方法相比,所提算法不仅保证了低动态载体速度计算的精确性,而且显著提高了载体速度在光照不均匀的室内环境下的计算效率,该研究为实现室内移动机器人实时视觉导航提供了理论依据。

A new method for fast calculation of low dynamic carrier velocity based on improved singular value decomposition (SVD)-Harris is proposed to solve the problem that the poor real-time performance of low dynamic carrier velocity calculation under the indoor environment with uneven illumination. Firstly, we use the SVD to compress and reconstruct the two adjacent visual images and use the improved Harris corner detection algorithm to detect the feature points of the two frames. Secondly, we use the normalized cross correlation (NCC) template matching algorithm to roughly match the feature points of two adjacent visual images. Thirdly, we use random sampling consistency (RANSAC) algorithm to eliminate the false matching point pairs. Finally, we use the information of the feature matching point pairs to calculate the carrier velocity. The experimental results show that the average calculation time of the traditional algorithm is 3.07 s, while that of the improved algorithm is 0.71 s. The error matching rate of the traditional algorithm is much greater than that of the improved algorithm. Compared with the traditional NCC template matching method, the proposed algorithm not only guarantees the accuracy of the velocity calculation of the low dynamic carrier, but also greatly improves the calculation efficiency of the carrier velocity under the indoor environment with uneven illumination. This study provides a theoretical basis for realizing the real-time visual navigation of indoor mobile robot.