摘要本文针对显微镜物镜焦深范围有限,显微样本成像时需要一序列图像才能聚焦清晰特点,对显微图像融合方法进行了研究,提出了一种基于拉普拉斯金字塔变换的显微图像融合方法,并初步实现了整套系统。63783
首先,本文用拉普拉斯金字塔变换融合算法对图像进行变换,分析这种图像融合算法的优缺点以及融合效果,并整理归纳了这种图像融合的主观评价方法。阐述了基于拉普拉斯金字塔变换的图像融合原理和方法:首先对源图像分别进行拉普拉斯金字塔分解,然后对分解后的各层图像采用不同的融合准则进行融合,最后,对融合金字塔做拉普拉斯金字塔反变换得到最终的融合图像。实验结果表明,该算法能得到具有更多有用信息的高对比度的融合图像,融合效果良好。
其次,在变换域融合算法上用 Laplacian 金字塔多聚焦图像融合算法。基于金字塔算法分解不带方向性,但变换存在数据冗余,且各层间数据存在相关性。对于低频系数的选取基于金字塔方法进行融合,仿真实验结果表明,该算法避免了重影,提高了图像的融合的质量。
毕业论文关键词 :显微序列图像,图像融合,拉普拉斯金字塔,聚焦合成
毕业设计说明书(论文)外文摘要
Title :Research and design of Pyramid Laplasse transform fusion based on microscopic images
Abstract
In this paper the microscope objective lens focal depth range of Co,microscopic sample imaging needs to a sequence of image focus clear characteristics, on the microscopic image fusion method is studied, presents a fast sequence image registration and fusion method, and realized the whole system.
First of all, the paper by Laplasse Pyramid transform is used to transform the image fusion algorithm,analysis of the advantages and disadvantages of this image fusion algorithm and the fusion effect, and summarize the various methods of evaluation of the image fusion. The paper describes the principle and method of image fusion based on the transformation of Laplasse Pyramid,the source images are decomposed Laplasse Pyramid, then the fusion of each layer of the decomposed image by using different fusion criterion. Finally, the fusion of Pyramid Pyramid Laplasse inverse transformation to obtain the final fusion image. The experimental results show that, the algorithm can get high contrast image fusion with more useful information.,good fusion effect.
Secondly, in the transform domain fusion of multifocus image fusion using Laplacian Pyramid. Pyramid algorithm based on decomposition without direction, but the transform data redundancy, and the data of each layer and the relationship between, and the data of each layer and the relationship between. For the low frequency coefficients are selected based on the method of integration of Pyramid, the simulation results show that, The algorithm avoids the ghost, improves the image fusion quality.
Keywords: Micro sequence image, image fusion, Laplasse Pyramid, focused synthetic
1 绪论 1
1.1课题研究的背景和意义 1
2 图像融合方法 4
2.1显微图像融合技术概要 4
2.1.1显微图像融合的目的 4
2.1.2显微图像融合的分类 4
2.2多聚焦图像融合算法 5
2.2.1多聚焦图像成像原理 5
2.2.2多聚焦图像融合方法概述 6
2.2.3空间域多聚焦图像融合