动态背景下基于 ORB 特征匹配的运动目标检测算法

摘要

In order to timely detect moving objects,we propose a moving object detection algorithm,which is based on the combination of ORB feature matching and the improved interframe difference.Firstly,we use ORB algorithm to extract feature points,and adopt RANSAC method to find the parameters of transformation matrix and then obtain the global motion compensation image.After that,we use interframe difference method to achieve the detection of moving targets.The high speed and accuracy of ORB feature point matching method as well as the effectiveness of RANSAC method in removing outliers ensure the accurate calculation of parameters of transformation matrix.Combined with interframe difference method,we rapidly and entirely detect the foreground objects.Experimental results show that the algorithm can accurately detect moving objects,and to some extent it can solve the issue of real-time detection.%为了能够实时地检测出运动目标，提出一个基于 ORB（Oriented FAST and Rotated BRIEF）特征匹配和改进的帧间差分相结合的检测算法。首先运用 ORB 算法提取特征点，采用 RANSAC（RANdom SAmple Consensus）方法求得变换矩阵参数之后获取全局运动补偿图像，然后用帧间差分法实现运动目标的检测。ORB 特征点匹配的快速性和准确性与 RANSAC 方法去除异常点的有效性确保变换矩阵参数的计算准确，再结合帧间差分法快速完整地检测出前景目标。实验结果显示，该算法能够准确地检测出运动目标，并在一定程度上解决了实时检测的问题。