文章基本信息

标题：A Comparative Study of Multi Scale Transform based Fusion Techniques
本地全文：下载
作者：Arwinder Kaur ; Ramanjeet Kaur ; Manish Mahajan 等
期刊名称：International Journal of Computer Science and Network
印刷版ISSN：2277-5420
出版年度：2016
卷号：5
期号：3
页码：429-436
出版社：IJCSN publisher
摘要：Image fusion is the progression of amalgamationof information from a number of cataloged images exclusiveof the introduction of deformation. Image fusion is ready tolessen the amount of data, maintain important informationand construct new image that is more appropriate foradditional processing tasks. High-resolution multispectralimages images can be formulated from a high-resolutionpanchromatic (PAN) image and low-resolution multispectral(MS) of remote sensed images using number of image fusionschemes. Multisensory fusion is used to achieve high spatialand spectral resolutions. It do so by merging images from twosensors, one of which has high spatial resolution and theother one has high spectral resolution in remote sensingregions. Image fusion technique is also used to eradicate noisefrom the images i.e. unwanted material which declines thequality of an image. Image fusion is applicable in many areaslike remote sensing, computer vision, military purposes andintelligent robots. The present paper analyses some of thesuccessful fusion levels such as pixel level, feature level anddecision level fusion. The paper also presents some of theimage fusion algorithms such as Average method, PCAfusion, high pass filtering which are spatial domain methodsand they deal with the image pixels directly. The temporaldomain fusion methods include transformation like DiscreteCosine Transform, Discrete wavelet transform. The methodspresented in the paper also have some influences and flaws
关键词：Fusion Levels; Principal Component Analysis;Intensity Hue Saturation; Multiscale transform; ANN; DCT