摘要针对单一传感器探测波段窄、提供信息种类单调了问题,本文提出了多源图像融合信息感知系统。该系统集成紫外光、可见光、近红外、中波红外和长波红外等五种成像探测器,通过图像融合技术,将不同波段采集的图像数据进行融合,舍去冗余信息,将多种探测器图像的有用信息综合到融合图像中去,以增强观察者对场景的理解,突出目标,提高对感兴趣目标的探测识别能力。本课题初步研究了多孔径图像融合系统异类图像配准技术、基于人眼视觉感知的图像融合算法以及显著目标融合检测算法三个关键技术问题,并对图像融合的算法进行了初步的研究,提出了利用基于拉普拉斯金字塔分解的图像融合算法来实现图像的融合。66376
毕业论文关键词:图像配准 图像融合 显著目标检测 拉普拉斯金字塔
毕业设计说明书(论文)中文摘要
毕业设计说明书(论文)外文摘要
Title Research in target extraction and identification method of Multi-source image fusion
Abstract
Because of the questions of narrow detecting band and boring type of information provision,this paper puts forward the information perception system of the multi-source image fusion. This system is integrated in ultraviolet, visible, near infrared, medium wave infrared and long wave infrared imaging detector. Through the image fusion technology,we can fuse image data collected in different wave bands an, giving up those redundant information and integrating those useful information of images from different detectors.In this way,we can enhance the understanding of the scene observers,highlight the objective and improve the ability of detection and recognition of interested target.This topic preliminary studied the image registration techniques of multiple-aperture image fusion system , image fusion algorithm based on human visual perception and target fusion detection algorithm significantly. Also, preliminary research was conducted on the algorithm of image fusion, proposing image fusion algorithm based on Laplacian pyramid decomposition is used to realize the image fusion.
Keywords Image registration;Image fusion; Laplace pyramid;Significant target detection
目 录
1 绪论 2
1.1 课题背景及意义 2
1.2 研究与发展现状 3
2 多源图像信息融合系统总体设计 4
2.1 现有图像融合系统研究存在的主要问题 4
2.2 系统的基本原理 4
2.3 系统构成 5
2.4 技术难点 5
3 多孔径图像融合系统异类图像配准技术 7
3.1 软件配准 7
3.2 硬件配准 7
4 基于人眼视觉感知的图像融合算法 9
4.1 基于拉普拉斯金字塔分解的图像融合算法的基本原理 9
4.2 实验仿真结果及分析 15
4.3 拉普拉斯金字塔融合算法的改进 20
4.4 小结 22
5 显著目标融合检测算法 23
5.1 显著性分析 23