一种基因芯片图像滤波混合法

对基因芯片的表达数据进行分析，有助于获得基因的表达谱与功能之间的关联信息，而基因芯片图像的滤波方法，对于获得高质量的基因表达数据具有重要的意义。本文采用小波和中值滤波混合法对基因芯片图像去噪，采用的小波硬门限阈值量化法去噪预先处理了基因芯片图像的部分高频噪声，避免了图像有用信号湮灭；而中值滤波法弥补了小波分析中的噪声指数较高的局限性。与传统的基因芯片图像滤波方法进行对比实验，结果表明，该方法能够在有效去除基因芯片图像噪声的同时，很好地保持图像的边缘和细节信息。

A method for denoising genechip images based on wavelet analysis and median filter

Genechip expression data analysis can reveal the relationship of gene profile and funcition.Denoising genechip image is very important to get veracious and reliable expression data.The method presented in this paper combines wavelet analysis and median filter to suppress the mixed noise of genechip images.Firstly,wavelet analysis filter is used for genechip image to filter the first level noise.It can wipe off part of noise while preserving edge and detailed signal information of image.Then median filter is used for the same image to wipe off the second level noise.The experimental results show that the method can effectively not only suppress noise but also preserve edge and detailed signal information of image compared with traditional methods.