Recovering high dynamic range radiance maps from photographs
Proceedings of the 24th annual conference on Computer graphics and interactive techniques
Radiometric CCD camera calibration and noise estimation
IEEE Transactions on Pattern Analysis and Machine Intelligence
True Images: A Calibration Technique to Reproduce Images as Recorded
ISM '06 Proceedings of the Eighth IEEE International Symposium on Multimedia
| Hi-index | 0.00 |
When recording presentations which include visualizations displayed on a monitor or with a video projector, the quality of the captured video suffers from color distortion and aliasing effects in the display area. A photometric calibration for the whole image can not compensate for these defects. In this paper, we present a per-pixel photometric calibration method that solves this problem. We measure the joint monitor-camera response function for every single camera pixel by displaying red, green, and blue screens at all brightness levels and capture them separately. These measurements are used to estimate the joint response function for every single pixel and all three color channels with the empirical model of response (EMoR). We apply the estimated response functions on subsequent captures of the display to calibrate them. Our method achieves a mean absolute error of about 0.66 brightness levels, averaged over all pixels of the image. The performance is also demonstrated with a calibration of a real captured photo, which is hardly distinguishable from the original.