图象处理在脉冲加热红外热成像定量检测中的应用

在红外热图象的分析研究中，由于噪声和热扩散效应，引起图象模糊和边缘扩散，影响了图象的进一步分析和面积大小的确定．为了改善图象质量，对本实验系统中得到的红外热图象进行了图象平均、中值滤波、二值化和梯度变换等处理．经过处理后的图象特征清楚，定量检测缺陷面积误差一般小于６％．

Application of Image Processing for Pulsed-heating Infrared Thermography Quantitative Testing

In the infrared thermography non destructive testing, image blurring introduced by noises and heat diffusion in the sample is an important problem, which influences the further analysis and the detestation of detecting area. In this paper, image averaging, median value, binarizition and gradient alternation are adopted in order to improve the quality of images. The processed images characteristics are clear, and the error of detesting area is smaller than 6%.