Design And Implementation Of Cartilage Slicing Image Stitching And Three-dimensional Reconstruction System

With the development of electronic technique, the research of electronic system which used to analyze the biological artificial material has been one of the main research areas in the medical electronic system development. Cartilage plays an important role of bearing load, reducing friction between bones in human body and so on. But for some unrecoverable damage, artificial cartilage is needed for replacement. In the cultivation process of artificial cartilage tissue, three-dimensional stress-tolerant ability of cartilage tissue is very important. Therefore, the simulation analysis platform of cartilage tissue has great application value. During the establishment of simulation platform, the cartilage tissue slicing image stitching and three-dimension (3D) reconstruction is one of the key technologies.In this paper, based on the relevant domestic and overseas document, the deep research has been done on image stitching and 3D reconstruction and designed a cartilage tissue slicing image stitching and 3D reconstruction system. This system contains two main functions:seamless stitching of partial slicing image and the 3D reconstruction and display of continuous slicing image.Before 3D reconstruction, the small view angle of lens dues to the high-resolution of electron microscopy, makes it impossible to capture an entire slice image, so this paper firstly focus on the image stitching of slice image. This paper use scale invariant feature transform (SIFT) algorithm to do the feature point detection and then do feature points matching for overlapping sections and the overlap region is determined by whether matching points are in the threshold, finally this paper realize the stitching function of two images. After that the paper shows the research on the 3D reconstruction of slicing image. Firstly, this paper enhance the images by reducing the noise, grey level transforming and so on and then do the 3D reconstruction. The implement of 3D reconstruction uses ray casting algorithm of volume rendering method. The ray casting algorithm uses the theory of light pass through translucent optical material, makes color composite for three-dimensional objects firstly, and then project it onto the window to output 3D reconstruction result. This 3D image can be displayed man-machine interactive also.This system works on Visual Studio 2010 platform, combined with OpenCV, ITK and VTK, using hybrid programming of C and C++ language to realize. The experimental data are divided into two types:Electron microscopic section digital image (SEM) and laser virtual slicing images. For image stitching function, use SEM to test, the reduction degree of image stitching is high and the accuracy of this system can reach 99%.For 3D reconstruction function, because of the poor data continuity of SEM, Laser virtual slicing images are used for the test. The 3D reconstructive result is good and it will be convenient for observation to do deep research.