Research On Color Consistency Of Remotely Sensed Tile Images

With the rapid development of remote sensing technology,the spatial resolution of remote sensing images is gradually improving,and the scale of remote sensing image data is also increasing.Compared with the traditional low spatial resolution remote sensing data,the information content of high-score data is significantly improved and the details of ground objects are more abundant.However,the application of high-score data is restricted by the huge amount and complexity of data.The Institute of Remote Sensing and Digital Earth of the Chinese Academy of Sciences has proposed an efficient data segmentation management method called the five-layer and fifteen-level tile segmentation method,which can play an important role in the efficient management of remote sensing image data and image production.Through data aggregation of remote sensing data of different sensors and different phases,the difference of data sources is effectively eliminated.However,in the process of satellite image acquisition,due to the difference of sensor types,shooting angle and time,and ground object attributes,there are some phenomena such as uneven brightness and color difference within a single image and between different images.Therefore,it is necessary to process the image color consistency,otherwise it will directly affect the seamless mosaic and real-time display of tile data.Based on the analysis of the existing uniform color algorithm,the uniform color algorithm between single tile and multiple tiles is systematically studied in this paper.The main research contents are as follows:(1)The existing single image smoothing algorithm and inter-image smoothing algorithm are studied.The existing algorithms are used to carry out experiments on different types of tile images,and the characteristics and applicability of each algorithm are analyzed..It is also pointed out that the MASK smoothing method is the best in the single image smoothing experiment,and Wallis filtering smoothing method is the best in the inter-image color consistency processing experiment,which provides help for subsequent research.(2)In the process of single image smoothing,the disadvantage of MASK smoothing method is that it cannot adopt the appropriate filtering scale for each region.Therefore,a dynamic block method based on image gradient is proposed to improve MASK smoothing method.The image blocks are classified by setting gradient threshold values,and the appropriate filtering scale is selected for each image block to carry out uniform light processing.The detailed information of the image is effectively maintained,and the quality of uniform light and shadow images is improved.(3)The Wallis filter color blending method based on basemap can not consider the positional relationship between pixels and can not improve the uneven brightness contrast in the original image.The color blending algorithm is improved to avoid blocking effect on the basis of consistent brightness contrast of images and to realize color consistency processing between images.