【作者】 张云峰； 赵建； 韩广良；

【Author】 Zhang Yunfeng Zhao Jian Hang Guangliang (Changchun Institute of Optics Fine Mechanics and Physics Chinese Academy of Sciences Changchun 130031 China)

【机构】 中国科学院长春光机与物理所；

【摘要】 针对运动目标的识别跟踪中要求同时满足实时性和准确率的问题,提出了一种基于边沿特征不变矩动态匹配的目标识别方法。首先,通过传统的SOBEL边沿提取方法获得完整清晰的目标轮廓,并基于该轮廓信息构造了一个具有平移、大小和旋转不变性的特征不变矩;然后提出动态极值匹配法,利用特征曲线的极值信息点进行识别匹配,并动态替换原特征模版。实验结果表明,该方法对目标运动姿态发生变化的情况具有很强的适应能力,利用较少的系统资源即可对运动目标进行实时跟踪识别。更多还原

【Abstract】 Objective to the requirement of both real-time and accuracy for moving target detect and tracking,a target recognition method is proposed based on dynamic matching of the moment invariants of edge feature. Firstly,acquire the more complete and clearer contour using Sobel mask.and based the contour constitute a feature invariant without influence of object’s size change,shifting,and rotation;then proposed a dynamic extremum-matching mehod,recognizing and matching the object based on extreme point information and replacing the former moldplate with the new one.The experimental results indicate that the method can successfully detect moving target and accurately estimate its trajectory in the image sequence with less memory even when the object figure is changed during movement,moreover,it is efficient and adaptable to real-time target detection and tracking.更多还原