Study On Microscopic Auto-Focus System Of Cell Screening Platform

Domestic study of cell screening platform is less than foreign countries in the field of biological cell culture,people generally adopts 96-well plates as the cultured medium to culture cell,owing to cells of community are smaller within the 96 orifice respectively,people's eyes could not recognize the cells,cell screening work needs to be observed under the microscope,cell selection work is facing that the screening detection rate is low artificial selection can make the operator feel fatigue,manual focus could be much time-consuming and so on.Aiming at these problems,the cell selection platform is designed,to study the platform of the cell screening microscopic auto-focus system,including to build cell screening platform,the design of motion control system,micro auto-focus system optimization of image preprocessing and auto-focus algorithm,on the premise of intact institutions,through the control unit of the driver,The carrier of inverted biological microscope has automatically achieved feed and step function by the control unit drive on cell screening platform.Secondly,depending on clarity value of a image taken by CCD,then transferring the image information to the auto-focus processing units,which achieve the objective of vertical feed,and then microscopic auto-focus system has been running.This paper has analyzed the working principle of the cell screening platform and the designing principle of microscopic auto-focus.To study the microscopic auto-focus window selection method in the system.In view of the traditional automatic focus area selection method to focus on cell microscopic image,due to the size and location of the focus area is fixed,and the cells belong to the random distribution,cells are not always in the center of the image area,resulting in the problem that while focusing,the main body cell community is usually ignored.So puts forward a automatic focusing window selection method based on Fish-Swarm algorithm by using the improved FishSwarm algorithm through increased the breeding and selection behavior,avoiding appearing the situation of the local optimal solution.Through the different images from the traditional focusing window selection method and Fish-swarm window selection algorithm,this paper respectively compares and analyzes the results from the Fish-swarm window selection algorithm and the improved Fish-swarm window selection algorithm.The results show that the image which is obtained by the improved Fish-swarm window selection algorithm can contain more details,and thealgorithm can automatically look for the main targets and then generate the focusing area;After using Fourier transform function to process the image for focusing areas,the number of high frequency is more and image has a higher resolution.Using traditional auto-focus gradient function evaluation algorithm on the image for microscopic auto-focusing,cells in the microscopic image for edge gradient gray value is smaller,the image is easily affected by noise when focusing.First,four direction operator template will be added to Sobel gradient function.According to the signal superposition principle,and the Sobel4 direction is weighted superposition on Brenner gradient function,which gets an improved Sobel gradient auto-focus function evaluation algorithm by changing degree of discrete data distribution,and it can improve the focusing accuracy,which is based on traditional gradient function and the improved gradient auto-focus function after the contrast experiment,the results show that the improved auto-focus evaluation algorithm can be better to restrain the noise,with a more traditional motor in mountain climbing algorithm for searching-focusing has smaller focus search range,and The degree of a image clarity value is also higher.The results in this paper illustrate the mechanism of cell screening platform,which reveal focusing accuracy of microscopic auto-focus in the system and the selection of focusing window law is of great significance,the results can be laid the foundation for recognition and classification of cell screening,which has an important application prospect in biological cell engineering.