Research And Application Of Biochip Image Gridding Algorithm

In the early 1990 s,proposed a new biotechnology was pioneered by gene chip.It is a highly intersecting new technology that covers electronics,biology,physics,chemistry,and computer science.Because it can to quantify the biological problems and the study of high flux is this technology is difficult to finish,so since it has caused the widespread attention at home and abroad,is widely used and clinical disease diagnosis,drug discovery and genome,and other fields.Biochip technology consists of five parts: sample acquisition,chip fabrication,scanning imaging,processing image and data analysis.Biochip image processing is an important part of biochip technology,including four steps: preprocessing,gridding,segmentation and information extraction.Therefore,our research on biochip is mainly research on its digital image processing.At present,many researchers have studied the two parts of grid and sample segmentation.Since only the location and shape of the sample points on the biochip image can be determined,the pixel with biological information can be extracted accurately and effectively.Can say gridding is the basis of image processing and data analysis.Because of the grid processing,we can determine the location of the sample points,which will greatly reduce the difficulty of sample segmentation and information extraction.At the same time,fully automatic grid processing is the key to the quantitative analysis of biochip images.Therefore,in this paper,we will study the grid algorithm based on the digital image of biochip,and we have done the following work:On the one hand,we have proposed a gridding algorithm which based on the Maximum Between-Class Varian(Otsu)and improved grid lines optimizing to solve the phenomenon of grid line redundancy,missing and appearing in the signal point.First calculate the image gray value projection in horizontal and vertical direction,combine with mathematical morphology reconstruction theory,find the optimal threshold which is obtained by using the maximize between-class variance;next we can get the binary reference signal of microarray image by the optimal threshold and determine the initial grid;then we should estimate the optimize the grid lines;last we find the final grid lines and draw them.The experimental results show that the accuracy of the proposed algorithm is significantly higher than other algorithms,and the algorithm is simple and feasible.On the other hand,the gridding system of reflected light images based on porous silicon microarray is designed.The system is using MATLAB graphical user interface(GUI)design,the system will contain a complete grid process.It contains the functions are: input the original image,obtain the important parameters,the important parameters display,gridding image display and log out.In this part,we will focus on describing the whole process of designing the graphical user interface and displaying the effect pictures,so that we can grasp the performance of the proposed algorithm from the whole.In general,MATLAB System Simulation based image gridding processing system of the biochip image powerful,simple user interface,easy to operate.