基于连续视差图差分算法的运动人体检测方法研究  

Research of Human Movement Detection Method Based on a Differential Algorithm of Continuous Parallax Figure

作　　者：张大禹[1] 黄灿[1] 

机构地区：[1]中国人民解放军92124部队

出　　处：《自动化技术与应用》2013年第5期36-39,共4页Techniques of Automation and Applications

摘　　要：运动人体的分析技术是计算机视觉技术的重要研究领域之一,主要研究内容包括运动人体的检测、跟踪、理解和描述。运动人体检测就是在视频序列中将运动人体区域从复杂的背景中提取出来。本论文采取了一种基于连续视差图差分算法的检测方法,即先估计相邻两帧的全局运动模型参数来对参考图像帧视差图做全局运动补偿,对相邻两帧视差图进行差分,再将差分图像进行二值化处理、形态学滤波和连通分量提取并去掉一些较小的连通区域,从而得到准确的运动人体区域。通过实验验证,该方法可以获得准确的运动人体检测结果,并在检测精度和检测速度之间取得了平衡。The analysis technology of the human body movement is an important research areas in the field of computer vision research, its main research contents include the movement of human detection, tracking, understanding and description. Detection means the body area extracting from the complex background in the video sequence. This paper uses a detection algorithm which is based on a continuous parallax figure of differential method, First of all, estimates the two of the global motion model parameters adjacent to the reference image frame disparity map global motion compensation. Adjacent pair of two parallax images are treated by the difference method, and the difference image of the binarizied procession, morphological filtering and connected components are treated to extract and remove the smaller communi- cating area, thereby obtains an accurate movement of body area. Verified by experiments, the method can obtain accurate test results of human movement and strikes a balance between the detection accuracy and detection speed.

关 键 词：运动人体检测 二值化 差分 滤波 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]