Radiometric Self Calibration

A simple algorithm is described that computes the radiometric responsefunction of an imaging system,from images of an arbitrar), scene taken using different exposures. The exposure is varied by changing either the aperture setting or the shutter speed. The algorithm does not requireprecise estimates of the exposuresused. Rough estimates of the ratios of the exposures (e.g. Fnumber settings on an inexpensive lens) are sufJicient for accurate recovery of the responsefunction as well as the actual exposure ratios. The computed response function is used tofuse the multiple images into a single high dynamic range radiance image. Robustness is tested using a variety of scenes and cameras as well as noisy synthetic images generated using IO0 randomly selected response curves. Automatic rejection of image areas that have large vignetting effects or temporal scene variations make the algorithm applicable to not just photographic but also video cameras. Codefor the algorithm and several results...

• Type: Conference
• Year: 1999
• Where: CVPR
• Authors: Tomoo Mitsunaga, Shree K. Nayar
• Keywords: Actual Exposure Ratios, Algorithm Applicable, Computed Response Function, Computer Vision, CVPR 1999, Simple Algorithm, Temporal Scene Variations