CAJ | 학술논문

为了改进传统三帧差分的运动目标检测方法易出现的空洞和与虚假边缘的现象，提出了一种基于五帧差分和边缘检测相融合的运动目标检测方法。该算法首先利用优化的 Canny 边缘检测算法快速提取五帧连续图像的边缘图像，然后对五帧连续的边缘图像进图像的五帧差分运算，最后通过阈值分割和形态学处理完成对目标的提取。该方法计算简单，实验结果表明，该算法准确率高，连通性好，能满足实时性检测的要求。
위료개진전통삼정차분적운동목표검측방법역출현적공동화여허가변연적현상，제출료일충기우오정차분화변연검측상융합적운동목표검측방법。해산법수선이용우화적 Canny 변연검측산법쾌속제취오정련속도상적변연도상，연후대오정련속적변연도상진도상적오정차분운산，최후통과역치분할화형태학처리완성대목표적제취。해방법계산간단，실험결과표명，해산법준학솔고，련통성호，능만족실시성검측적요구。
To improve the traditional three frame difference of moving target detection method in which the hole edge and false edge tend to appear, this paper puts forward a method based on the fusion of five frame difference and edge detection, moving object detection method. The algorithm firstly uses Canny edge detection algorithm of optimized fast extraction of five consecutive images edge image, and then to the edge of the frame of five consecutive image into five frames difference operation. Finally, by threshold segmentation and morphological processing is completed on the extraction of target. The method is simple in calculation and the experimental results show that the algorithm has high accuracy, good connectivity, and can meet the requirements of real-time detection.