摘要
手术中医生的头、手等部位运动产生的阴影直接影响了手术的质量和病人的健康,为了检测出这些运动遮挡物,进而消除阴影,文中提出一种基于改进的高斯混合模型的运动物体检测方法。该方法从更符合人眼视觉特性的HSV色彩空间对图像进行建模并提取运动前景,改善了原算法性能;结合HSV、RGB色彩空间中的像素特征和差分算子进行阴影抑制,充分利用了阴影与运动物体的色彩信息;针对阴影边缘产生的干扰采用形态学操作消除,提高了检测结果的准确性。实验证明,本方法能准确地提取出运动物体,抑制噪声干扰,为进一步研究手术区域灯光补偿工作奠定基础。
Since the shadow caused by the motion of doctor＇s head, hands and other parts is directly related to the quality of surgery and patient＇s health in surgery. In order to effectively detect the moving occlude and apply it for shadow removal, a detection method of mov ing target based on improved Gaussian mixture model （ GMM} is proposed. The GMM is built in HSV color space which is fit for human visual feature, improving its performance. The HSV and RGB color space feature and difference operator are considered to shadow sup- pression which is make full use of the color information of shadow and moving target. Finally, remove the disturbance of shadow by using morphological operation ,improve the accuracy of detection result. The experimental result shows that this method can effectively detect the moving target, restrain noise interference and make foundation for further study of shadow removal.
出处
《计算机技术与发展》
2012年第12期139-141,145,共4页
Computer Technology and Development
基金
5D数字式LED手术无影灯及其工作方法专利(CN101858537A)
关键词
运动物体
高斯模型
阴影抑制
HSV色彩空间
moving target
Gaussian mixture model
shadow suppression
HSV color space
