The Research Of Uniform Color Method Based On Remote Sensing Images

With the development of remote sensing technology,the technique of uniform color and mosaic of multi view remote sensing images is one of the image processing technologies which have been developed rapidly in recent years.Using multi view remote sensing images to carry on uniform color and mosaic can produce a panoramic image which is more complete,comprehensive and reliable to view of the earth.But whether the remote sensing images obtained by the way of aerial remote sensing or UAV airborne radar,due to weather conditions,human factors and sensor factors,remote sensing image are inconsistent in color and brightness,resulting in that causing big error when they used in image mosaic or it will has the obvious color difference image mosaic in the two sides of splice joint on the mosaicked image,affecting the visual effect of entire map.So before the remote sensing images used for image mosaic,they require to uniform color with an image as the reference image,but there is not a perfect uniform color method to solve all influencing factors when remote sensing images are used in the field of the image mosaic.On the basis of the remote sensing image processing,t his paper studies around the uniform color method of remote sensing image s.This paper firstly introduces two kinds of classic uniform color processing method,which named uniform color based on histogram matching processing method and uniform color based on Wallis filter processing method.Secondly,on the basis of the classical uniform color algorithms which do not deal with the details,this paper presents the uniform color method based on bilateral filter.And bilateral filter is used for the remote sensing images processing in this paper.Finally,it introduces the uniform color algorithm proposed in this paper,which named the uniform color method based on combining the global with local information.This algorithm is suggested on the basis of the remote sensing images that often appear several color and contrast problems in the uniform color processing.The uniform color algorithm proposed in this paper can be divided into three parts.The first part is the color model processing,in which makes the images convert color space.It is aimed to minimize the correlation between each channel of images,in order to process individually in each channel.The second part deals the overall image brightness.According to the spatial entropy of the image luminance component,it can calculate the spatial histogram.And a whole image processing mapping function can be obtained according to the spatial histogram.Then the whole image of the luminance component can be processed.The third part is the local image brightness processing.In this part,it is the further processing of image brightness components after whole image processing,and then the brightness components treated by two dimensional discrete cosine transform.The weighted operation carried out in the image transform domain.The luminance component can remain after local processing,and the detail part is treated effectively.Through experimental simulation of measured data,the uniform color processing algorithm proposed by this paper based on combining the global with local information can well solve the color problems of images and achieve good visual effect,close to the real color,clear texture and so all.