Algorithm Study On Layers Association Based Slice Contour Smoothing

Medical three-dimensional reconstruction is a very active advanced research field, and it is the outcome of computer and medical technology. The three-dimensional object model reconstructed by the human two-dimensional slice images has been widely applied in the medical and educational fields, and it is of great realistic significance. Reconstruct three-dimensional objects from the three-dimensional slice images has a wide range of applications in many other areas.Two-dimensional images is the basis of three-dimensional reconstruction, therefore, the quality of it will directly affect the results of 3D reconstruction. Based on investigation and analysis of the two-dimensional slice contours and the actual results of three-dimensional reconstruction, this paper presents a smooth algorithm on the human slice contours before the three-dimensional reconstruction. The algorithm combines the contour smooth based on the B-spline fitting according to the slice correspondence information with the smooth in one layer, and using the character points collected from the section sampling model, all the slice contours can be smoothed effectively. From the result of three-dimensional reconstruction based on the smoothing contours, we can see that the saw tooth drops and the visual effects improve significantly. The algorithm is universal, and it is also applicable to other areas of contour data smooth.In order to understand the algorithm deeply, this paper also gives a contour smooth algorithms based on the Bezier surface fitting, and then makes a detail analysis of the distinction and respective advantages between the two algorithms. In order to improve the visual effects further more, the author also implements the Laplace algorithm to smooth the triangulated surfaces after reconstruction, which gives a more profound understanding of smooth algorithm before and after three-dimensional reconstruction.