| China is the world’s most important pork producer and consumer,but with the outbreak of African swine fever in recent years,the pork farming industry has been hit hard,seriously affecting pork production.Increasing the number of pigs by improving the success rate of breeding pigs is an important means of increasing production.The detection of sperm motility of breeding pigs is an important part of the pork breeding process.Screening out sperm with strong motility can increase the success rate of breeding.At present,there are two common ways to detect sperm motility,namely manual observation by microscope and computer-assisted semen analysis(CASA).However,these two detection methods have the following problems:(1)The manual observation method of the microscope is judged by the technicians observing under the microscope with experience.This method relies on subjective experience,high randomness,low accuracy,and lack of data support;(2)Although the emergence of CASA has greatly improved the accuracy of semen analysis and made sperm motility detection standardized,however,CASA needs to rely on bulky microscopes and computers,which greatly limits the scenarios of sperm motility detection.In view of the above problems,this paper proposes a sperm motility detection method based on a microfluidic chip combined with a smartphone platform.The main work is as follows:(1)In order to simulate the movement process of sperm in natural environment,a bionic microfluidic chip based on sperm natural optimization is designed.The detection is more in line with the real sperm movement environment;(2)Three types of classical target detection algorithms are studied,and then the advantages and disadvantages of each algorithm are analyzed,and the improved adaptive Gaussian mixture model algorithm is used to separate the foreground and background of sperm according to the characteristics of sperm,and segment motile sperm targets to prepare for the subsequent tracking process;(3)A variety of target tracking algorithms are studied,and the applicable scenarios and constraints of each algorithm are analyzed.Using the Kalman filter algorithm to predict and track sperm targets,and use the IOUbased Hungarian matching algorithm to perform correlation matching of multiple sperm targets between adjacent frames to achieve sperm multi-target tracking.According to the movement trajectory of the sperm target,the kinematic parameters of the sperm are calculated,and finally the motility of the sperm is classified according to the parameters;(4)A miniature microscopic imaging device is designed and fabricated,which can be paired with a microfluidic chip and a smartphone to form a portable sperm motility detection platform.The sperm motility testing APP based on Android platform was developed,which can replace the traditional microscope with computer.After testing,the portability detection platform designed by the project captures clear images,and the shooting video is stable,which is close to the image and video quality of traditional acquisition methods.The sperm motility detection algorithm used has a recognition rate of 96.47% and an average tracking rate of 88.55%,which can effectively identify and track motile sperm targets and calculate sperm motility.The designed microfluidic chip can effectively filter low-motility spermatozoa and realize the natural preferred function.In conclusion,the sperm motility detection method based on the microfluidic chip combined with the smartphone platform proposed in this paper can replace the detection method of traditional microscope and computer,and is easy to operate,highly portable,and is not limited by the usage scenario. |
PDF Full Text Request