Font Size: a A A

Research On Multi-Camera 360-Degree Panoramic Image Mosaicing Technology

Posted on:2016-02-27Degree:MasterType:Thesis
Country:ChinaCandidate:L J YingFull Text:PDF
GTID:2308330479993982Subject:Systems Engineering
Abstract/Summary:PDF Full Text Request
In the field of video surveillance, increasing the angle of view of the surveillance cameras become a critical perspective and need of technology.More than in this context,this paper uses the camera hardware platform,focuses on the implementation of 360-degree panoramic image stitching technology. the paper is based on the predecessors’ research, improved on the existing algorithm and implemented the process, to meet at the same time high quality image stitching and video real-time requirements.In view of the panoramic video stitching algorithms requiring real-time and robustness, the paper carrys on the important research and improvement in the following several aspects.1, The paper,to meet 360- degree panoramic stitching, needs to be redesigned for hardware platform. on the basis of the original four cameras, it is added a camera, to increase the size of image overlap area. on the other hand, it can increase the calculation time,,so the whole system need to optimize the algorithm.2, In this paper, on the basis of the original brightness adjustment,,we add the ajustment of vignette,to adjust the average brightness difference of the whole image, especalliy the differences of brightness values among the brightness edges and center, making the result of the whole image mosaicing more good.3, In the original image distortion correction, the process is manually adjust the distortion center and distortion coefficient, generate mapping table, but the speed of the method is too slow, moreover the result is very poor, so the paper to improve the original part of distortion correction, is automatically calculating the distortion center and distortion coefficient using the module.4, Analysis the advantages and disadvantages of feature matching and pixel matching, the paper finally uses Harris-SSD matching at the beginning of feature points matching, reducing the search scope,then uses correlation matching,the result of matching is remarkablely improving.5, In order to overcome the difference of brightness and mild parallax on the transition zone,the paper combine of the multiband blending and the best juncture blending on the fusion of the original calculation, to improve the quality of the blending.
Keywords/Search Tags:multi-camera system, Wide angle lens, Distortion correction, Image registration, Image blending
PDF Full Text Request
Related items