Improving sperms detection and counting using single Gaussian background subtraction

摘要

Research about determining infertility rate of sperm is still being under constant development. First important phase on the sperm infertility observation is detection of sperm object. Success rate of separate sperm with semen fluids has important role for further analytical measure. This research is on its ways to detect and count human's spermatozoa. Detected sperms are moving sperm that is moving on the video. To detect moving sperm, Single Gaussian background subtraction is used. This method fits for sperm detection because the sperm data used are tends to be in unimodal. This research also uses other methods of background subtraction as comparison. The examination result shows that Single Gaussian method has f-measure value 0.853 and successfully extracts the sperm shape fully better than other methods.