Feature Estimation of Circular Metal Blooms using Image Processing Approach

摘要

This paper, presents a circle detection algorithm and its implementation to estimate diameter of circular objects such as metal bloom. Bloom is a mass of iron, steel, or other metal rolled into a thick bar in circular shape for further work. Work image consisting is subtracted from reference image to get required region of interest [ROI]. ROI is pre-processed to remove the noise using median filter. Further filtered image is then converted into binary image for further processing. The next processing steps include detection and diameter calculation of metal blooms kept in yard. The proposed algorithm gives good results and consumes less memory space and minimal computation time as compared Hough Transform [HT] based circle detection algorithm.