关键词: 砂轮堵塞, 红外检测, 热像图, 图像分割, 区域生长分割法

Abstract:

According to the grinding wheel loading detection problems, a method was proposed for detecting the grinding wheel loading based on the active infrared detection technology and the thermogram segmentation technology. Three methods including the threshold segmentation, the edge detection and the region growing were used to segment the thermogram, the features like the amounts and states of the grinding wheel loading were extracted. The experimental results show that: the distribution of the grinding wheel loading swarf extracted by the threshold segmentation method presents a “segregation” phenomenon, and the selected threshold value is affected the loading features deeply. The edge detection method has overcome the “segregation” phenomenon, but there still exist some problems such as the edge is not complete and the positioning of the edge is not accurate. The results of the RGS method based on the gradient threshold filtering are more objective and accurate. The test result errors of the grinding wheel loading detection between the RGS method and the microscopic photography were given. The RGS method can realize active quantitative detection of the grinding wheel loading.

Key words: grinding wheel loading, infrared detection, thermogram, image segmentation；region growing segmentation(RGS) method