一种基于图像金字塔光流的特征跟踪方法

An Image Pyramid-Based Feature Detection and Tracking Algorithm

推导并实现了一种基于图像金字塔光流的角点特征跟踪方法。实验结果表明,该方法在不同运动幅度和运动方式下的检测跟踪性能较好,能够有效地应用于长序列图像的特征跟踪。

An image pyramid-based feature detection and tracking algorithm is dedaced and implemented. The experimental results prove that it performs well when subject to different magnitudes and types of motion, it is more robust and can be effectively applied in feature detection and tracking through long image sequences.