The Fusion Research Of Distance Informaiton And High-resoluiton Color Image

Distance information is one of the very popular research direction in the field ofcomputer vision, PMD (Potonic Mixer Device) camera can get range information of targeteffectively in real-time. However, PMD camera still exist limitations, the random noise isbig and the horizontal resolution is low. This thesis fuses the distance information taking byPMD camera with high resolution color image to obtain high resolution color3D data, andto do the3D object reconstruction in viewing area. The research work of this thesis has abroad application prospect and practical value.In this thesis, the main work is as follows:(1) Introduces three types of data getting by PMD camera and the measuringprinciple of the camera, analyzes the advantages and the existing limitations of thecamera. Summarizes the domestic and foreign related research on PMD camera work,and the existing PMD camera distance and high resolution color image fusion methods.Designs the2D/3D camera imaging system.(2) Proposes PMD distance information and RGB image data fusion based onbicubic interpolation with credibility. The method fully considers the relevance ofcredibility between the amplitude information and the distance information getting byPMD camera, which enhances the precise distance information’s impact on the fusionresults, and gives the2D/3D registration method using the distance information, whichreduces the complexity of the algorithm.(3)Since the PMD distance information and RGB image data fusion based on doublethree interpolation with credibility hasn’t highlighted the edge, we proposes PMDdistance information and RGB image data fusion based on edge information, which keepsthe edge information of fused images effectively. And according to the fusion method,we put forward a distance image quality evaluation method based on edge strength.