3D meshes with vertex colors can be produced using single RGB-D sensor with motion tracking. Consumer grade cameras do not always automatically support HDR video mode, but store RGB images in color space such as sRGB which is optimized for single exposure and gain value, When fusing color information from multiple views with different gain and exposure values, raw pixel measurements need to be mapped into HDR color space which allows generating photographs with arbitrary exposure and gain values to match individual photographs. In this work, a stack of images was stored using different ISO and exposure values in typical room lighting environment with bright windows. A mathematical model was found by experiment which models mapping from camera parameters into pixel values. Post-process gain and exposure control was tested using the output HDR 3D model.
HDR1
HDR2
HDR3
HDR4