基于灰度共生矩阵和区域生长算法的红外光伏图像分割

随着光伏产业的迅速发展，这种光伏新能源在各个方面正扮演越来越重要的作用，因此对光伏面板的维护和状态监测就变得尤为急切。而对光伏面板红外图像的分割是后期对光伏面板故障检测的基础。因此，针对红外光伏图像对比度低、信噪比低等特点，本文提出一种改进的区域生长算法结合灰度共生矩阵(GLCM)的纹理特征和图像的梯度特征对红外光伏面板图像进行分割，该算法通过图像的梯度特征和灰度共生矩阵的熵特征对原始红外光伏图像进行预处理，获得梯度特征和熵特征图像加权叠加后的预处理图像，然后再利用红外图像灰度直方图分布集中的特点，针对需要人工干预进行种子点的选取，实现自动进行种子点的选取，同时对区域生长准则进行改进，使其能够自动的调整生长阈值，从而可以分割出光伏面板区域。实验结果表明，本文算法相比OTSU和K-means聚类算法分割效果更好且更接近手动分割的目标区域。

Infrared Photoelectric Image Segmentation Based on GLCM and Region Growing Algorithm

With the rapid development of the photovoltaic industry, this new photovoltaic energy is playing an increasingly important role in all aspects, so the maintenance and condition monitoring of photovoltaic panels becomes more urgent. The segmentation of the infrared image of the photovoltaic panel is the basis for the later detection of the failure of the photovoltaic panel. Therefore, in view of the low contrast and low signal-to-noise ratio of infrared photovoltaic images, this paper proposes an improved region growing algorithm combined with the texture features of the gray level co-occurrence matrix(GLCM) and the gradient features of the image to segment the infrared photovoltaic panel image. The gradient features and the entropy characteristics of the gray level co-occurrence matrix preprocess the original infrared PV image, obtain the pre-processed image after the gradient feature and the entropy feature image weighted superposition, and then use the characteristics of the infrared image gray histogram distribution concentration, Manual intervention is needed to select the seed points, and the seed point selection is automatically performed. At the same time, the regional growth criterion is improved, so that the growth threshold can be automatically adjusted, so that the photovoltaic panel area can be segmented. The experimental results show that the proposed algorithm is better than the OTSU and K-means clustering algorithms and is closer to the target segmentation.