基于自适应混合高斯的改进三帧差分算法

Improved three-frame difference algorithm based on adaptive mixture Gaussian

为提高运动目标检测算法的准确性,保证较低的时间复杂度,提出基于自适应混合高斯的改进三帧差分算法。为获取目标内部运动点,采用基于自适应学习率的混合高斯背景建模,以像素点间的匹配次数作为参考量来修正模型的学习速率,提高算法对动态环境的适应性,通过基于边缘提取的三帧差分改进算法对视频图像的目标轮廓进行提取,保证目标信息的完整。实验结果表明,所提算法能够完整提取运动目标,保证目标轮廓的完整,算法的时间复杂度有效降低。

To improve the accuracy of the moving object detection algorithm,ensure lower time complexity,an improved three-frame difference algorithm based on adaptive Gaussian mixture was proposed.To obtain the internal movement points of the target,a Gaussian mixture background modeling based on adaptive learning rate was used,and the number of matching between pixels was used as a reference to modify the model learning to improve the adaptability of the algorithm in the dynamic environment.Through the three-frame difference improvement algorithm based on edge extraction,the target contour of the video image was extracted to ensure the integrity of the target information.Experimental results show that the proposed algorithm can completely extract moving targets and ensure the integrity of the target contour.The time complexity of the algorithm is effectively reduced.