Research Of Microscopic Image Acquisition Control System And Image Sparse Denoising

The information which micro vessels images contained can reflect theorganization form of pathological changes, it can provide fundamental supportfor disease diagnosis, prevention and recovery through the analysis of the microvessels image information. But during acquisition process of micro vesselsimages, the physical activities of live animals lead to micro vessels of interestdeviates from the area that camera can focus on, and the heart beat of liveanimals can blur the images acquired by camera. What’s more, the existence ofoptical and electronic noise can cut down the equality of micro vessels imagesin the process of image acquisition.In order to solve the problem that micro vessels of interest deviates fromthe area that camera can focus on due to body activities of live animals, and thechange of distance between live animals and the camera caused by heartbeatbrings about the low clearness of image, a small live animals microscopicimage acquisition platform is designed and realized under MFC in VisualStudio2012. When the live animals deviates from the area that camera canfocus on, the platform move to the opposite direction of the small animal inorder to make the micro vessels of interest back into the area that camera canfocus on at once.Then, the movement control of platform was designed and realized. Thecontrol of micro vessels images of the interest moving into the area that cameracould focus on, is achieved in real time and accurately, to avoid the microvessels image of the interest deviating from the area that camera could focus onagain through adjustment manually or switch control. The real-time control andaccurate position of camera is achieved to improve the clearness of microvessels images, and solve the effect of optical noise on micro vessels images in certain degree.Finally, the micro vessels images is denoised based on the sparserepresentation method. A dictionary suitable for the acquired micro vesselsimage is trained by K-SVD algorithm based on the characteristics of images,and then the image after the process of denoising is acquired through the sparsedecomposition algorithm of OMP. Compared with the original noisy image, thequality of images after denosing based on sparse representation is improved alot. Comparatively speaking, the image denosing method based on sparserepresentation proposed in this article is adaptive. This method can train adictionary suitable for the image to be denosied for the different images.