由于成像特点及环境干扰,用于工件缺陷检测的红外热像通常较为模糊.为此,将空间域与频域结合的模糊C均值(FCM)聚类算.法用于红外热像中缺陷及正常表面的分割.运用多种图像处理方法对原始红外热像进行预处理,将得到的高频图像及其邻域平均图像使用经典FCM聚类算法进行像素灰度的聚类.实验结果表明,该方法的分割效果较好.%Imaging principle and environment disturbance have lowered the definition of the infrared thermal image. Aiming at this problem, an infrared thermal image processing framework is proposed based on Fuzzy C-means(FCM) clustering algorithm with spatial and frequency information. A sequence of preprocessing is applied to the original infrared thermal image. FCM with spatial and frequency information is used on the preprocessed image to segment defect and normal surface of a part. Experimental results show that the proposed infrared thermal image processing framework is very effective to detect the surface defect of a metal part.
展开▼