COMBINED METHOD OF RED BLOOD CELLS COUNTING ON BLOOD SMEAR IMAGES
Read the full article
The problem of red blood cells counting on images obtained by microscope often appears within computer-based analysis of blood products. Segmentation method of those images is presented in the paper, and it is based on Canny edge detector with following circle identifying with Hough transform. Segmentation results refinement with binary mask of objects and background is used in the article. The proposed method is resistant to structure, texture and blood cells color changing, and also it allows detecting cells that overlap each other even on noisy images.