|
Image interpolation plays an important role in three-dimensional reconstruction, and the interpolation results have a great impact on the reconstruction directly. A feature-distance-based region matching interpolation method is presented in this paper: First, the corresponding relation between the interpolated images is built by an optimization process which is based on the gray value, the gradient of the gray value, the gradient direction in neighboring slices and the gray-scale continuous threshold which is set to get the paired points. Second, for the best-matching point pair, linear interpolation, and for the worst-matching point pair, we calculate the shortest feature distance to the relative feature points and use the shortest feature distance to control the image interpolation. The results show that the new method does not require the image shape information and can improve the image fuzzy border problem of internal organs and enhance the quality of the interpolated images. |
|
Keywords:image interpolation;region matching;threshold;relative feature distance |
|