THE RESEARCH OF MEDICAL MICROSCOPIC IMAGE MOSAIC BASED ON THE ALGORITHM OF SURF

Microscopic diagnosis of the prostate is one of the methods commonly used in clinical medicine. The small sheet of the acquired microscopic medical images gives rise to some problems in clinical diagnosis. To solve it, we research microscopic image mosaic correspondingly. In this paper, the SURF algorithm is used to complete microscopic image feature detection and description, then we combine Bray-Curtis distance and k-d tree algorithm to complete "coarse matching". then these obtained matching points are used to achieve "fine matching" by RANSAC algorithm, finally fade in-out algorithm is used to perform image fusion, microscopic image mosaic is achieved lastly. Experiments show that its feature extraction is accurate for slices of microscopic image of the prostate cancer, the result is promising so that it has a good assisted effect for diagnosis of the disease.