A New In-Camera Imaging Model for Color Computer Vision


We present a study of the in-camera image processing through an extensive analysis of more than 10,000 images from over 30 cameras. The goal of this work is to investigate if image values can be transformed to physically meaningful values, and if so, when and how this can be done. From our analysis, we found a major limitation of the imaging model employed in conventional radiometric calibration methods and propose a new in-camera imaging model that fits well with today's cameras. With the new model, we present associated calibration procedures that allow us to convert an sRGB images back to their original CCD RAW responses in a manner that is significantly more accurate than any existing methods. Additionally, we show how this new imaging model can be used to build an image correction application that converts an sRGB input image captured with the wrong camera settings to an sRGB output image that would have been recorded under the correct settings of a specific camera.



In this section, we provide more experimental results.

sRGB to RAW: Here, we present more results not shown in the paper of converting nonlinear sRGB images to RAW responses for various scenes and cameras as explained in the paper.Please click links below to see the examples:

Photo refinish: Please see the introduction and results video here. For more results and convenient comparison, please check the examples from Canon&Nikon and Sony&others.

Data sets

Please check the following links to download the available datasets:

canon sets nikon sets other sets
Canon sets Nikon sets Other sets

Code/Interface: The code/interface for exploring the data sets can be downloaded from here:

download code brief instruction
Code in matlab(source and .exe) Brief instruction

* About how we extract the colors from JPG/RAW images, please download the preprocessing code in C++ and an example of using it.

Last updated: 30 Dec 2011