Research on Segmentation of Protozoan and Metazoan Image in Microscopic Examination of Activated Sludge

摘要

Monitoring water quality using biological monitoring technology based on microscopic examination of activated sludge, machine vision replaces manual observation. The contrast between target and background is relatively low in the images obtained by automatic electron microscope. It is difficult to accurately distinguish the target and background if single image segmentation method is used. Therefore, the image which is segmented by single image segmentation method will be regarded as the source image, and image fusion is performed on these source images using wavelet transform. The fused image will be regarded as the final segmentation image. The result of experiments shows that using this idea can get better image edge detection effect, and it will provide good technical basis for the next protozoan and metazoan identification and statistics.