Automatic human spermatozoa detection in microscopic video streams based on OpenCV

The analysis of semen plays an important role in male fertility evaluations. Computer-aided Sperm Analysis systems have been working on providing more accurate information about the sperm motility and quantity. However, the existing sperm detection algorithms which segment sperms according to grey levels are not able to preclude bright non-sperm objects, like the round cells. The contribution of this paper is a solution to this problem. The use of Gaussian-modeling method makes our algorithm able to filter the bright non-target objects. We also apply the morphological image processing method to our algorithm to improve the targets dispersion quality. This algorithm has good prospects and accomplishes both an accuracy of 95% in average and a real-time processing according to our tests. It is helpful for semen quality analysis, eugenics and test-tube babies.