基于模糊C均值聚类和Canny算子的红外图像边缘识别与缺陷定量检测

针对脉冲红外热成像检测缺陷构件时，红外图像噪声较大、边缘信息模糊等特点，提出了一种基于模糊C均值聚类和Canny算子相结合的边缘检测新方法。该方法首先对输入的红外图像进行整体灰度变换，采用模糊C均值聚类对图像进行区域分割、提取和二值化；再将各个区域进行叠加，使红外图像的边缘变得连续；最后，采用Canny算子对处理后的图像进行边缘检测，实现缺陷的识别。在图像边缘检测基础上，分析了图像定位缺陷位置与实际缺陷位置之间的相对误差，并运用物像关系，实现缺陷几何尺寸的定量检测。结果表明:该方法对缺陷边缘识别完整清晰，具有较高的定位精度和抗噪能力，有利于缺陷的识别与定量检测。

Infrared image edge recognition and defect quantitative determination based on the algorithm of fuzzy C-means clustering and Canny operator

A new edge detection method based on Fuzzy C-means clustering and Canny operator was proposed to detect the defects of infrared thermal imaging with large noise, edge information ambiguity and so on. In this method, the gray scale transformation of the input infrared image was carried out, and the image was segmented, extraced, and binarizated by the Fuzzy C-means clustering; then each area was superimposed to make the edge of infrared image continuous. Finally, the image was processed by the Canny algorithm, and the edge of the infrared image was continuous. Canny operator was used to detect the edge of the image, and the defect recognition was realized. Based on the image edge detection, the relative error between calculated and actual defects position was analyzed, and the geometric size quantitative detection of defects was realized. The results show that the proposed method can detect the defect edge completely and clearly, and has higher accuracy and anti-noise ability, which is advantageous for the identification and quantitative detection of defects.