Study On Pixel-level Multi-source Remote Sensing Data Fusion Technique

With the double-quick development of the sensor,the remote sensing platform and data communication technology, the multi-source remote sensing images data(multi-sensor,multi-platform,multi-spatial resolution,multi-spectral images,multi-temporal) of the same region are more and more which are accessed by all kinds of sensor. But, different types of RS images have different applying characteristic and objects and there are corresponding limitations also. Multi-source remote sensing images data fusion is the one of most effective way to enrich these multiple sensor remote sensing information. As a technology of remote sensing information's comprehensive treatment and analysis, the research of multi-source remote sensing image fusion is becoming an important research direction in the remote sensing field. Based on analyzing and summarizing the existing document and researching result, from studying pixel-level fusion technique and it's application and taking panchromatic images of high spatial resolution and multi-spectral images of low spatial resolution as data source, this paper carries out the following research work :1,Introducing the basic concept of multi-source remote sensing data fusion and analyzing and summarizing the aim,mode,level,flow and method of multi-source remote sensing information data fusion.2,To ensure remote sensing images quality, this paper studies several kinds of processing means.3,This paper enunciates the basic concept of multi-source remote sensing image pixel-level fusion and it's characteristic. It introduces six fusion methods detailedly such as product fusion method,ratio fusion method,high-pass filtering fusion method in spatial domain and IHS transform fusion method,Principal Component Analysis fusion method> wavelet-transform fusion method in transform domain. It offers scientific evidence for choosing the remote sensing information fusion method. Based on analyzing and summarizing the existing methods of assess the image quality, the combination of subjective qualitative evaluation and objective quantitative evaluation criterions is the effective method to assess fused images quality and of fused images methods.4,This paper gives some result by testing four common arithmetic and using criterions to assess test result. By analyzing test result, this paper sums up the characteristic and range of application of six common fusion methods.