| 以下仅列出文章摘要、提纲简介，如需获取全文阅读权限，或原创定制、长期合作，请随时联系。 微信QQ:312050216 |
 |
 |
| 基于L0最小化稀疏重建与梯度求导模型对斑马鱼图像的去模糊处理 摘 要： 斑马鱼是现今常用的模式生物之一，有着很重要的研究价值。因为定量研究需要，对经过药物处理的斑马鱼图像清晰度要求较高，但是由于种种客观原因常造成图像模糊，并且模糊核无法提前得知。关于图像盲去模糊的问题，本文提出一种基于L0最小化稀疏重建与梯度求导模型对斑马鱼图像去模糊处理，利用引入中间变量进行梯度求导求得逼近L0最小化的解。同时，分别使用Tenengrad梯度法、Laplacian梯度法和Variance法对本文以及其他算法对斑马鱼图像去模糊实验结果进行图像清晰度评价，经过比对，本文算法在三种评价指标下均取得最优结果，说明了算法的有效性。 关键词：图像去模糊；稀疏重建；斑马鱼；梯度求导 中图分类号：TN919.8 文献标志码：A 文章编号： Zebrafish image deblurring based on L0 minimization sparse reconstruction and gradient derivation model YANG Weijie, XU Jianyu (Institute of electrical engineering and computer science, Ningbo University, Ningbo 315211, China) Abstract: Zebrafish could be one of the most commonly used model organisms, because it has important research value. Because of the need of quantitative study, the clarity of the zebrafish after drug treatment is high, but due to various objective reasons, the image is often blurred and the fuzzy kernel can not be known in advance. In view of the classic problem of image blind de-blurring, this paper presents a deblurring algorithm for zebrafish image based on the L0 minimization sparse reconstruction and gradient derivation model. By introducing the intermediate variable into taking derivatIve of the gradient , the solution to minimize L0 is proposed. At the same time, Tenengrad gradient method, Laplacian gradient method and Variance method were used respectively to evaluate the image sharpness of zebrafish image defuzzification experiment using this algorithm and other algorithms. After comparison, the proposed algorithm achieves the best results among the three evaluation indexes, and The effectiveness of the algorithm is illustrated. Key words: image deblurring; Sparse reconstruction;zebrafish; Gradient derivation  |
