论文部分内容阅读
2代Bandlet变换中耗时最多的是四叉树构建和最优几何方向选取,而且原算法准确度也不高.为了解决这个问题,本文提出基于进退法选取最优几何方向算法;然后又提出了基于遗传算法的最优方向选取算法;最后把这两种方法结合在一起可以取得更好的结果.同时改进了原算法中的自底向上的四叉树算法为自顶向下的四叉树算法,提高了对特定图象处理的速度.最后本文把改进的算法应用到去噪,压缩和图像增强中,取得了较好的效果.并对算法的时间和空间复杂度进行了分析.
In the second generation of Bandlet transform, the most time-consuming is quadtree construction and the selection of the optimal geometry direction, and the accuracy of the original algorithm is not high. In order to solve this problem, this paper proposes to select the optimal geometric direction algorithm based on the advance and retreat method; The optimal direction selection algorithm based on genetic algorithm is proposed.Finally, the two methods can be combined to get better results.At the same time, the bottom-up quadtree algorithm in the original algorithm is improved to the top-down quadruple Tree algorithm to improve the speed of the specific image processing.Finally, the improved algorithm is applied to denoising, compression and image enhancement, and achieved good results.And the time and space complexity of the algorithm is analyzed.