基于MATLAB的基因芯片图像预处理

目的：基于MATLAB设计低密度基因芯片图像预处理系统，对经过Scan Array扫描系统获得的用cy3和cy5染色的低密度芯片的荧光图像进行处理，以滤除噪声。增强图像的对比度，提高图像的质量，实现图像的边缘检测与分割，进行区域的识别。方法：采用小波中值滤波方法，滤除噪声，提高图像的质量；基于小波方法，结合边缘算子进行图像边缘检测；使用遗传算法实现图像的分割。结果：该系统对基因芯片图像进行处理，减少了污点、噪声及其他各种因素的影响，提高了图像的质量，可以更好地检测出图像的边缘，比较准确地分割出样点区域，能有效地分离有价值的弱信号点和背景点或者噪声。结论：该系统可以基本实现低密度基因芯片图像预处理功能．采用的图像处理方法是可行的，能为以后的分析提供较为准确的数据信息。

Image Pre-processing of Gene Chips Based on MATLAB

Objective To design an image pre-processing system of low-density gene chips based on MATLAB, which can process the colored images by cy3 and cy5 fluorescence staining of low-density chips obtained by the array scanning system. It can filter out noise, enhance the contrast gradient of image, improve the quality of image, and implement the functions of image segmentation, edge detection and region identification. Methods The median filter method of the wavelet was used to implement the function of image denoising and improve the image quality. Edge detection was accomplished by wavelet, combining with edge operators. Image segmentation was developed by genetic algorithms. Results It could reduce the effect of spot, noise and other factors, improve the quality of image, and detect the periphery of image better and the region of sampling point more precisely. It can also effectively separate the valuable weak signal points and background or noise with the system. Conclusion The system can accomplish the functions of image pre-processing of low-density gene chips, and the adopted methods are feasible. It can provide relative accurate data information for future analysis.