Research Of Medical Microscopic Image Mosaic Based On SURF Algorithm

Image mosaic is a method that joints several images which have some overlap into a complete large view image. It has been widely used in many fields, especially has important application value in medical microscopic image mosaic. At present, prostate cancer incidence remain at a high level, biopsy is the only way to diagnose. Because of the microscope vision is too small, especially at high magnification, overall information become very difficult to get sliced. Therefore, this paper carries out some research on this issue with the aid of image mosaic technology, which can splice automaticly and effectively.In this paper, it uses "SURF-BC" algorithm. Firstly, the method uses SURF algorithm to complete the feature detection and description of medical microscopic image.Then, to calculate the similarity of feature points between the images base on the Bray-Curtis distance, in order to finish the coarse matching between the images. Next, further eliminates false matching points by RANSAC algorithm to achieve the fine matching between the images. Finally uses the method of gradually into the fade out on microscopic image fusion, achieving microscopic image mosaic.In this paper, Gleason4prostate acinar adenocarcinoma microscopic images are used to test and verify "SURF-BC" image mosaic algorithm. Based on the above theory, experiments show that, microscopic images seamless stitching can be achieved by the method, and the corresponding experimental results is very promising. Furthermore, a system model is achieved in the use of Matlab software programming, which users can complete the work of medical microscopic image mosaic by clicking on the corresponding menu.