Design Of A Cell Counting Method And Its Detection Device

Cell counting is a basic and important work in the fields of biology,cytology and medicine.cell counting is to detect the number of cells in a specific volume of cell samples.and take it as an important means to evaluate the ability of cell proliferation and the health status of biological individuals.therefore,it is of great significance to realize the automation of cell counting.In this paper,we design and develop a fully automatic microscopic cell counting system with autofocus,which is mainly used in the bright field environment,and realizes the detection of cells in the bright field environment.Automated counting of cells in suspension.The cell counting system is mainly composed of core parts such as imaging system and software system.The imaging system is mainly responsible for collecting clear cell images and transmitting them to the computer.The imaging system uses a CMOS industrial camera combined with a low-magnification microscope objective to image and collect cell images,and uses an edge detection operator as the autofocus evaluation function.Combined with the hill-climbing search algorithm,the automatic focusing function of image acquisition is realized.The software system is responsible for the control of the stepper motor and the realization of the cell counting algorithm.In the process of cell imaging,there will be a small amount of cell stratification.Due to the limited depth of field of the imaging system,cells with different layers in suspension cannot be clearly collected by industrial cameras.Therefore,these cells will not be counted and the counting result will not be smaller than the actual result.Aiming at the phenomenon of cell stratification,this paper uses an image fusion method based on pulse coupled neural network to fuse cell images with different faces,which can improve the accuracy of cell counting.Cell counting algorithm is the core part of this paper.This paper proposes a cell counting method based on the segmentation of cell center highlight region.Combined with the characteristics of flood filling algorithm,this algorithm segments the center highlight region of suspended cells as the identification of living cells,and obtains the number of cells by counting the number of connected regions in the binary image.In addition,an image processing algorithm for identifying dead cells and cell clusters is proposed.Experiments show that the accuracy of the algorithm for counting the number of cells in the cell image is more than 99%,which can meet the requirements of cell counting in the laboratory.On this basis,this paper also studies the cell counting method based on density estimation,using the center point of the segmented cell center highlight area as the cell labeling point to make a cell image number set,and filter the cell image.Get the correspondence between pixels and density values in the image.The convolutional neural network is used to return the cell density map on the cell image,and the number of cells is counted by finding the local maximum value in the cell density map.In this paper,a complete set of cell counting and cell imaging device design scheme is proposed,which provides a new fully automatic cell counting method for relevant researchers in the process of cell culture.