Microarray Image Detection And Analysis System Based On Digital Image Processing

Microarrays(biochips)have always been a very important new technology in the field of life sciences.At present,a fluorescent substance is used for labeling in a commercial microarray test.And the microarray is analyzed by the detected fluorescence intensity.The fluorescent markers are too expensive,and may cause changes in the structure and activity of biomolecules.To solve these problems,a new type of porous silicon microarray was proposed.In this paper,we apply digital image processing technology to study the characteristics of this new porous silicon microarray image,and put forward our own method and opinion.The key technologies in microarray image processing include image preprocessing,tilt correction,sample segmentation,speckle denoising,and gray value calculation.The main research contents are as follows:Firstly,the processing of porous silicon microarray images is introduced.The binarization of images is an important part of the microarray image processing.It directly affects the accuracy of subsequent processing such as tilt correction and sample segmentation.Based on the characteristics of noise in porous silicon microarray images and some common methods,the preprocessing method of HSV space vector combination is proposed in this paper.The superiority of the method is verified by experiments.Secondly,it is difficult to avoid the scanning deviation for the scanning instrument to obtain the image,which causes the image tilt and causes trouble for the subsequent image processing.Therefore,according to the geometrical characteristics of porous silicon microarray images,two algorithms which are simple,accurate and fast are proposed,and the image's tilt correction is realized.Thirdly,the sample segmentation is in order to determine the area which needs to be measured.It is a key step in the data extraction of the porous silicon microarray image.On the basis of summarizing the commonly used segmentation methods and combining the unique characteristics of porous silicon microarray images,a segmentation algorithm based on array cells is proposed.This method can eliminate the bright spot on the right side of the array cell effectively,and ensure that the desired sample points are segmented in the effective cell area,and the accuracy of subsequent gray value calculation is improved.Fourthly,due to the roughness of the porous silicon surface,the scattering light interferes with each other after the laser incident.The speckle noise will be formed on the photosensitive surface of CCD.The presence of speckle noise has a great influence on the gray value of the porous silicon microarray cell,which leads to the inaccurate result of the final calculation.Therefore,the influence of speckle noise on gray value is analyzed,and a method to reduce the influence of speckle noise on the gray value of porous silicon microarray image is proposed,which improves the accuracy of detection results.Finally,an image processing and analysis system for porous silicon microarrays is developed.The system can complete tilt correction,sample segmentation,speckle denoising,and calculate the average gray value of all points after segmentation.So it achieve a complete set of porous silicon microarray image detection and analysis process.