Orthophoto Seamline Network Automatic Search Method

As one of the 4D products,Digital Orthophoto Map(DOM)has the advantages of high precision,rich information,intuitive and realistic,and quick access.It has a wide range of applications in emergency disaster relief,measurement and mapping,and land and engineering development.Because it is difficult for a single image to cover a target area,acquiring multiple DOM images by tessellation is an important step in the process of remote sensing image processing.In this paper,the automatic search method for inlaid lines is studied.One of the key techniques for differential correction of image mosaics is to search for a reasonable mosaic line within the image overlap area.Images on both sides of this mosaic line do not produce misalignment and differences in brightness and chromaticity.Because brightness and color difference can be removed by uniform light evenness,and if the mosaic image has a geometrical misalignment,it will make obvious stitching seams near the inlaid line.This is even light uniformity can not be removed,will greatly affect the image Splicing quality.Therefore,how to avoid geometrical misalignment in the mosaic process is a key issue in the mosaic process.The general inlaid line search algorithm searches for an optimal path on the cost image,but the algorithm for finding the optimal path is difficult to maintain the balance between the global optimum and the time complexity of the algorithm.Globally optimal algorithms(such as Dijkstra's algorithm)tend to have high time complexity,and algorithms with low time complexity can only guarantee local optimality.For the problems in the inlay line search process,the main research work in this paper is as follows:(1)The reason of the geometrical misalignment caused by orthoimage mosaicing is analyzed.A method of describing geometrical misalignment using disparity maps or dense optical flow diagrams is proposed.The SGM dense matching algorithm is used to obtain the orthophoto map disparity maps or the Farneback algorithm is used to extract the denseness.The optical flow diagram is used as the cost image.When the mosaic line advances along the low-luminance area of the cost image,the mosaic image of the mosaic line will have the smallest geometrical misalignment.(2)An automatic search method for orthographic mosaic lines based on the principle of minimizing the maximum edge weight is proposed.Firstly,a disparity map or a dense optical flow map is used as a cost image,and a disparity image and a difference image are superimposed to generate a difference image,which is regarded as a weighted undirected graph.Then,the best mosaic is searched based on the Bottleneck model using the minimum maximum edge weight algorithm.Line,and use block strategy to further optimize to improve operating efficiency.(3)Experiments show that the proposed algorithm can effectively avoid the geometric differences or the areas with large differences in chromaticity on the images,and obtain mosaic results that can meet the application requirements.