The 3D Model Acquisition Pipeline

摘要: Three-dimensional (3D) image acquisition systems are rapidly becoming more affordable, especially based on commodity electronic cameras. At the same time, personal computers with graphics hardware capable of displaying complex 3D models also inexpensive enough to be available a large population. As result, there is potentially an opportunity consider new virtual reality applications as diverse cultural heritage and retail sales that will allow people view realistic objects home computers. Although many physical techniques for acquiring data—including laser scanners, structured light time-of-flight—there basic pipeline operations taking acquired data producing usable numerical model. We look at fundamental problems range registration, line-of-sight errors, mesh integration, surface detail color, texture mapping. In area registration we both finding initial global alignment using manual automatic means, refining this variations Iterative Closest Point methods. To account scanner errors compare several averaging approaches. single joining from all scans, various methods computing interpolating approximating surfaces. then ways in which properties such color (more properly, spectral reflectance) can extracted imagery. Finally, examine final model representation efficiently rendered hardware.