| 基于图切割和阴影抑制的目标检测算法 |
| |
| 引用本文： | 李芬兰,李超.基于图切割和阴影抑制的目标检测算法[J].汕头大学学报(自然科学版),2009,24(2):56-62. |
| |
| 作者姓名： | 李芬兰 李超 |
| |
| 作者单位： | 汕头大学工学院电子系,广东,汕头,515063 |
| |
| 摘 要： | 针对目标检测中背景噪声影响检测精度的问题,提出在HSV高斯背景模型下,使用最小图切割算法,并结合连通性,找到最大的前景连接区域,最后使用阴影抑制技术完成目标分割.实验结果表明,算法有很好的鲁棒性,能够获得清晰、光滑的前景目标轮廓.
|
| 关 键 词： | 图切割 阴影抑制 目标检测 高斯背景模型 |
Object Detection Based on Graph Cuts and Shadow Elimination |
| |
| LI Fen-lan,LI Chao.Object Detection Based on Graph Cuts and Shadow Elimination[J].Journal of Shantou University(Natural Science Edition),2009,24(2):56-62. |
| |
| Authors: | LI Fen-lan LI Chao |
| |
| Institution: | (Department of Electronic Engineering, Shantou University, Shantou 515063, Guangdong, China) |
| |
| Abstract: | In order to solve the problem of detection precision, which is affected by the background noise, an algorithm which is based on the HSV Gaussian background model is proposed, and the minimum graph cut algorithm is used to segment objects. It is combined with the pixels connectivity to find the largest connected component. The shadow eliminating method is used to finish the segmentation. Experimental results show the robustness of the method. Clearer and smoother contour of the objects are obtained. |
| |
| Keywords: | graph cuts shadow elimination foreground detection Gaussian backgroundmodel |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
