A method of moving objeet detection based on image sequences is proposed in this article. The phase correlation algorithm is adopted to calculate the image background rotation and translation caused by the camera's movement, and then the adaptive threshold method is introduced to perform the smooth processing on the image and image registration processing. The high pass filtering is used before differential multiplication to remove the noise and enhance the image edge for extract the moving objects. The good performance and adaptability of the method were verified by the experiments.%提出一种基于图像序列的运动目标检测算法.采用相位相关算法计算由于摄像机运动而导致的图像背景的平移量和旋转量,引入自适应阈值对待处理图像做平滑处理,进行图像配准处理,最后在差分相乘算法之前采用高通滤波处理,滤除噪声,增强边界以提取运动目标.实验结果表明该方法在工程实践中具有良好的性能和很强的适应能力.
展开▼