Improving Selectivity on High-Temperature Decarburization Depth Measurements Using an Image Segmentation Method

Optical decarburization measurements are preferred for quality tests, modeling, or indirect methods over carbon steel products. The measurements are commonly done by metallographs and can be supported by image processing. Image processing allows the collection of multiple measurements from digital micrographs. Nevertheless, limited attention has been paid to observing the influence of the segmentation methods on the precision of the depth measurements. The present paper uses type 1045 steel micrographs to examine the impact of the segmentation method parameters on the precision of depth measurements obtained with image processing. The micrographs are processed by a median filter and statistical region merging process. Then, a script is utilized to obtain 2,560 measurements per micrograph. The data is used to depict visual results of the decarburized zones and detailed statistical analysis. Eventually, image processing and manual measurements are compared to denote the reproducibility of the method. The paper concludes that the segmentation parameter can be used to measure complete or total decarburized zones for further analysis.