基于Sobel算子边缘检测的太阳电池缺陷特征提取方法

RESEARCH ON SOLAR CELLS DEFECTS FEATURE EXTRACTION BASED ON SOBEL OPERATOR EDGE DETECTION

摘要: 利用太阳电池的红外图像对其进行缺陷的特征提取研究,提出基于Sobel算子边缘检测的太阳电池缺陷特征提取方法。首先,在RGB空间和HSV空间中,对太阳电池的红外图像进行预处理,得到太阳电池的R(红)、G(绿)、B(蓝)灰度图和H(色调)、S(饱和度)、V(明度)灰度图,初步判断太阳电池的运行状态;其次,使用Sobel算子对非正常运行的太阳电池红外图像进行边缘检测,得到红外电池表面增强图像的Sobel细节图,并利用自适应迭代法获得最佳阈值,进一步得到有较强表示能力的阈值突显边缘图,充分体现了太阳电池表面的缺陷特征细节。

Abstract: By using the infrared thermal image of solar cells,a method of solar cells’ defects feature extraction based on Sobel operator is presented. First,the infrared thermal image of solar cells is pre-processed to obtain R（red）,G（green）,B（blue）grayscales and H（hue）,S（saturation）,V（value）grayscales in RGB and HSV spaces,and the working status of solar cells is prejudged. Next,the edge information of solar cells’ infrared thermal image in abnormal operation are extracted by using Sobel operators,also,the adaptive iteration is employed to obtain a threshold highlighting edge graph with strong representation ability.