摘要
去污作业时,工程人员需要准确定位放射性区域,彻底去除有害物质。提出了一种利用图像匹配定位放射性区域的方法,通过普通相机即可确定污染物位置。具体过程为:首先,采集放射性区域的γ相机图像,并提取辐射成像区域的边缘轮廓;然后,从γ相机和普通相机图像中找出匹配点云,并利用点云计算出相机间的变换矩阵;最后,把γ相机图像中的放射性区域轮廓映射到普通相机图像中,完成普通相机对放射性区域的重新定位。实验过程中:γ相机与普通相机有水平和垂直偏角时,该方法通过普通相机准确找到了模拟的放射性区域;而在应用现场,该方法同样通过普通相机找到了放射性污染物。
During decontamination,engineers need to accurately position radioactive area and remove hazardous material. A radioactive area positioning method based on image matching is proposed, and the contaminant can be found with normal camera. There are three steps ： firstly, the γ camera image is acquired and edge contour of radioactive area is extracted; secondly, from the gamma camera image and normal camera image, the matching point cloud is got and the transformed matrix is formulated ; lastly, the contour is mapped into normal camera image and radioactive area is relocated by it. In experiments：when there are some horizontal or vertical deflection angles, this method can accurately find out simulation radioactive area; at application field, it also accurately position contaminant with normal camera.
出处
《电视技术》
北大核心
2016年第5期110-115,共6页
Video Engineering
基金
核能开发科研项目(2011-1137)
四川省教育厅重点项目(14ZA0091)
四川省科技支撑计划项目(2015GZ0035)
关键词
Γ相机
放射性区域
图像匹配
匹配点云
gamma camera
radioactive area
image matching
match point cloud
作者简介
肖宇峰（1978-），副研究员，主要从事计算机视觉与网络通信研究。
