利用红外成像技术检测绝缘子污秽等级的关键在于获取分类性能优异的红外热像特征,文章提出了基于方差分析的红外热像特征选择方法。利用直方图均衡化增强原始热像图对比度,人工截取绝缘子盘面图像区域;通过平滑后的图像直方图包络线提取分割阈值,对阈值分割后的二值图像进行形态学滤波,得到绝缘子盘面图像和背景图像,提取2者的最高温度、最低温度、平均温度、温度分布方差以及盘面相对于背景的最大温升和平均温升共10个红外热像特征;应用单因素方差分析甄别特征优劣,实现特征选择。瓷绝缘子人工污秽试验结果表明:文中提出的红外热像特征选择方法和图像分割算法简单有效。 The key of using infrared imaging technology to detect the contamination level of insulators is to obtain the infrared thermography features of excellent classification performance. In this paper, a method of feature selection based on variance analysis is proposed. Histogram equalization is used to enhance the contrast of the original thermal image, and the insulator disk image area is manually intercepted; the threshold is extracted from the smoothed image histogram envelope to morphologically filter the threshold-separated binary image to obtain the insulator disk image And the background image, the maximum temperature, the minimum temperature, the average temperature, the temperature distribution variance and the maximum temperature rise and the average temperature rise of the disk relative to the background were extracted. The infrared thermal imaging characteristics of the two were extracted. One-way analysis of variance , To achieve feature selection. The results of artificial contamination tests on porcelain insulators show that the method of feature selection and image segmentation proposed in this paper are simple and effective.