| 一种适合于多目标检测的图像分割方法 |
| |
| 引用本文： | 陈忠碧,张启衡. 一种适合于多目标检测的图像分割方法[J]. 光电工程, 2004, 31(5): 34-37 |
| |
| 作者姓名： | 陈忠碧 张启衡 |
| |
| 作者单位： | 中国科学院光电技术研究所,四川,成都,610209;中国科学院光电技术研究所,四川,成都,610209 |
| |
| 基金项目： | 国家863高技术项目资助 |
| |
| 摘 要： | 提出一种适合于多目标检测的8邻域图像分割方法,用以获取目标特征量,准确分割图像。在对二值化图像扫描形成目标块的过程中,标记各个目标像素,记录目标的边界点,得出分割目标的面积、周长、质心坐标。利用这些信息,可以选择跟踪一个或多个目标。仿真结果证明了该方法实用可行。
|
| 关 键 词： | 图像分割 二值化图像 边缘提取 图像处理 目标跟踪 |
| 文章编号： | 1003-501X(2004)05-0034-04 |
| 收稿时间： | 2004-02-10 |
| 修稿时间： | 2004-02-10 |
An image segmentation algorithm suitable for multi-object detection |
| |
| CHEN Zhong-bi,ZHANG Qi-heng. An image segmentation algorithm suitable for multi-object detection[J]. Opto-Electronic Engineering, 2004, 31(5): 34-37 |
| |
| Authors: | CHEN Zhong-bi ZHANG Qi-heng |
| |
| Abstract: | An 8-adjacent image segmentation algorithm that is suitable for multi-object detection is proposed. This algorithm will be applied to acquiring object features and segmenting image accurately. During scanning binary image to form object region, each object pixels will be labeled and object boundary points will be recorded, thus the area, circumference and centroid coordinate of the segmented object can be obtained. One or more objects can be tracked at one's option by using these data. The simulation results demonstrate that the method is practical and feasible. |
| |
| Keywords: | Image segmentation Binary image Contour extraction Image processing Target tracking |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
