论文部分内容阅读
提出一种新的图像多目标特征点集检测技术．该技术研究了多目标分割、目标细做形状提取和宏形状提取，进而检测表示目标案形状特征的特征点集．各目标案形状特征点集进行识别与跟踪．研究了上述技术的快速实现．实际DeP跟踪系统处理结果表明：该技术计算量小，抗噪能力强且准确性高，能准确识别目标，适用于实际图像目标特征点集的检测．
A new image multi-target feature point detection technology is proposed. The technique studies multi-target segmentation, target thinning shape extraction and macro shape extraction, and then detects feature point sets representing the shape features of the target case. The target case shape feature point set to identify and track. Studied the rapid realization of the above technology. The actual DeP tracking system processing results show that the proposed method has the advantages of small computation, strong anti-noise ability and high accuracy, and can accurately identify the target. It is suitable for the detection of the target feature point set of the actual image.