多聚焦图像融合算法的研究-应用文.docVIP

分类号：******** U D C：******-***-(20**)****-0 密 级：公 开 编 号：******** ************大学 学位论文 多聚焦图像融合算法研究 论文作者姓名： 申请学位专业： 申请学位类别： 指导教师姓名（职称）： 论文提交日期：  多聚焦图像融合算法研究 摘 要 光学成像系统焦距确定后，只有成像在景深区间内的空间点能够清晰成像。实际应用中对某个场景的物体成像时，由于被照场景中各物体与成像镜头的物距各异，所成的像不是全都清晰的。为了获取清晰的全场景图像，需要对场景中不同的物体分别聚焦，获取每个物体的图像，并将其融合在一起，即多聚焦图像融合技术。研究了多聚焦图像融合的基本理论，特别是空间域和变换域的融合方法；阐述了有关小波变换（WT）的融合理论，该办法通过对原始图像实施小波分解，将解析后的低、高频区域作相应的变换，高低频区域分别使用不同的融合规则，然后用修正后的小波子区域融合成新图像。设计了计算机模拟实验，对几种基于小波的多聚焦图像融合算法进行了模拟并给出了结果评价，实验结果证实了文中方法的有效性。 关键词：多聚焦图像融合；小波变换；图像重构；质量评价；融合规则The research on multi-focus image fusion algorithm Abstract After determining the focal length of the optical imaging system, only when imaging in the space point of the depth of focus can be clearly imaged. In the real process of the image-forming, because of the difference of the object distance between the things and imaging lens in the scene which be focused, the image-forming to certain scene is not all clear. To obtain clear panorama, we can respectively focus on the different objects in the scene, get all the images of the object and mix them together, this is so-called multi-focus image fusion technology. The basic theory of multi-focus image fusion, especially the spatial domain and transform domain fusion method; Expounded about the wavelet transform(WT) fusion theory, the approach, the original image by wavelet differentiation, will be resolved after the low-frequency region for the corresponding conversion, high-frequency regions are using different fusion rules, then corrected wavelet sub-regional integration into a new image. Design of computer simulation experiments, several fusion algorithm based on wavelet multi-focus images are simulated and gives the results of the evaluation, experimental results confirmed the validity of the method. Key words: multi-focus image fusion; wavelet transform; image reconstruction; quality evaluation; fusion rule 目 录 论文总页数：33页 1 引言 1 1.1 课题研究背景及意义 1 1.2 图像融合技术的研究现状及问题