Generation methods in 3D GIS of large Angle camera monitoring scene

Project number: 41101406

Main contents of the research:

The road monitoring camera joining together of bigAngle monitor scene in 3D GIS display is beneficial to the overall situationperception and evaluation. First of all, in view of the current lack of space layout design method, it is difficult to generate more optimization monitoring camera configuration scheme. This project which makes use of 3D GIS analysis function, combine with camera imaging model and the traditional static image splicing algorithm, create more than a factor evaluation system, put forward ageneral surveillance cameras space layout model. Secondly, making operation on the scale normalization, image edge enhancement to meet the video sequence image mosaic efficiency and adaptability requirement, put forward a kind of new large field view video image splicing algorithm, based on fast Fourier transform image allocation algorithm and calculating the two adjacent video frame between coefficient of registration, the improved method of linear systemof registration system for global optimization. Finally, the article proposes a kind of many camera video frame in the fusion processing algorithm, based on Walllis transform image pixel mean linear correction method adjacent video image color balance, eliminate coincidence area "ghost" phenomenon.This project will provide 3D GIS in camera video sequence of big Angle monitor scene generation to provide a new way and method.