摘要
该文采用改进最大类间方差(OTSU)法和区域生长法处理变电站巡检机器人在不同光照条件下所拍摄的仪表图像。针对于变电站工作环境的复杂性及其他不确定等因素,通过传统OTSU法将所拍摄仪表图像分成三类进行处理:正常光照条件、光照强度过亮、光照强度过暗,在正常光照条件下,直接采用传统的OTSU法进行二值化处理,能够正确分割目标区域与背景区域。在光照过暗条件下,在传统OTSU方法基础上进行改进,对仪表图像进行两次二值化迭代处理,对目标区域和背景区域进行两次分割处理,提取仪表图像感兴趣区域。在光照过亮条件下,采用区域生长法,对仪表图像进行分割处理。通过对巡检机器人所拍摄大量的图像进行验证分析,证实了所述方法的有效性、正确性。
In this paper, the maximum inter class variance method and the region growing method are used to deal with the instrument image that inspected by the substation robot under different illumination conditions.Because the substation working environment is complexed and other adverse factors.The instrument image is processed by the traditional maximum inter class variance.The image is divided three categories：Normal illumination conditions, light intensity is too bright, light intensity is too dark.Under normal illumination conditions.The target area and background area is correctly splited by used OTSU method. In the light of the dark,by improving the method of maximum inter class variance,the instrument image is processed two times binarization iterative.The target area and background area are segmented twice to extracting the interest area of instrument image. In the light of the light,the image is segmented by the region growing method.Through analysised a large number of images inspected by the robot,the method are validity and correctness.
出处
《电子世界》
2015年第20期145-149,共5页
Electronics World
