基于运动目标检测和骨架细化的算法研究

Research on Algorithm Based on Moving Target Detection and Skeleton Thinning

针对目前已有的运动目标检测算法和骨架细化算法的常见问题,提出改进的基于高斯混合模型(Gaussian Mixture Model,GMM)的背景减除法和改进ZS细化算法。传统的运动目标检测算法存在动态背景变化、光照变化、对噪声和阴影敏感等问题,笔者提出的改进的基于GMM的背景减除法可以较好地应对运动目标检测的挑战。该方法不仅可以减少光照变化的影响,降低噪声和阴影,还可以处理自然场景的动态变化。细化技术作为图像处理领域中最重要的技术之一,被应用于逐层侵蚀一个物体的图像,直到留下一个骨架。笔者对几种具有代表性的并行细化算法进行研究,并对目前应用最广泛的ZS细化算法进行改进。经实验验证,该改进算法可以克服ZS细化算法的缺点,得到拓扑性、连通性及细化性更强的人体骨架。

Aiming at the common problems of existing moving target detection algorithm and skeleton thinning algorithm, an improved background subtraction method based on GMM and an improved ZS thinning algorithm are proposed. Traditional moving object detection algorithms exist dynamic background changes, illumination, is sensitive to noise and shadow problems, this paper propose an improved background subtraction based on GMM division can cope with the challenges of moving target detection well, this method can not only reduce the influence of illumination change, reduce noise and shadow, also can deal with the dynamic changes of natural scenes. As one of the most important techniques in the field of image processing, thinning technique is applied to erode the image of an object layer-by-layer until a skeleton is left. In this paper, several famous parallel thinning algorithms are studied,and ZS thinning algorithm, which is the most widely used at present, is improved. Experimental verification shows that the improved algorithm can overcome the shortcomings of ZS thinning algorithm and obtain a human skeleton with stronger topology, connectivity and refinement.