The Design And Implementation Of Sperm Movement Detection System Based On OpenCV

With the development of society,the infertility has presented a rising incidence in younger trend.In the modern male subjects and male family planning,semen analysis has become an important test of male fertility.Traditional semen test methods are based on doctors observing the semen samples under the microscope.These methods are subjective and inefficient.Computer assisted semen analysis(CASA)is the typical application of multiple target tracking technology in the medical field.The system based on computers,image processing and machine vision which can detect and evaluate the sperm movement ability and morphology efficiently and accurately has a high value of clinical and scientific research.This is also a valuable research subject in image processing.Combined with the research of computer assisted semen analysis system at home and abroad,this project is based on the powerful image processing functions of OpenCV and has designed a complete sperm movement detection system.The system connects the camera at the top of microscope and controls it to film semen sample videos.By analyzing the motion situation of sperms in the video,the system makes an evaluation of activity of the sperm sample.Different sampling environment can result different sperm images under the microscope.In order to guarantee the consistency and accuracy,image enhancement,image smoothing and denoising are used for preprocessing.The relative operation contains histogram equalization,filtering and morphological operation.These operations are used for improving image quality,enhancing the contrast of background and targets,highlighting the sperm targets for post-processing.In this paper,different preprocessing methods are used for sperm images of biological microscope and phase contrast microscope to obtain the same effect for sperm motion detection.In the processing of sperm motion detection,overlapping averaging is used for obtaining the static targets and background subtraction is used for separating the dynamic targets.This paper implements the accurate tracking of sperm movement by applying the matching principle and the minimum displacement difference between minimum neighborhood method in dynamic target tracking process.At the same time,the paper improves the nearest neighborhood method,narrows the scope of target search,and adds a matching failure treatment to improve the accuracy of the matching results.This subject has implemented a complete sperm motion detection system based on Microsoft Visual Studio 2010 embedded OpenCV to display sperm motion visually and evaluate sperm motility detailly.The system has realized following functions:filming the sperm videos under the biological microscope and phase contrast microscope,detecting the sperm motion and boundaries,detecting the amounts of living and dead sperms,calculating relative sperm motion parameters based on WHO Laboratory Manual for the Examination and Processing of Human Semen(Fifth Edition)and so on.The further development and application of the system will add the sperm morphology detection for more complete automation of sperm quality evaluation system.This system has great advantages of reducing manpower and improving work efficiency.The research of this paper has a high reference value in clinical and medical research.