Color Correction for Projected Image on Colored Screen Using Correction Matrix Based on Camera

Recently, projectors are one of the most common display device for using presentation at office and class, entertainment at home, and theater. The use of mobile projector expanded the applications such as meeting at field and presentation on the spot. The images are not always projected on white screen, causing color distortion. Various algorithms have been proposed to correct the projected color the same as that on white screen. However, previous algorithms have limitation such as real-time performance due to high computational complexity and use of measurement equipment as well as lack of accuracy due to the use of matrix obtained by simple linear method. Accordingly, color correction method using a camera is proposed to match the color between on white and colored-screen. This method has advantage at the accurate color reproduction on coloredscreen by using ramp patches and regression method. It is differ from previous method using only two or three patches. In the proposed method, ramp patches for each channel are first projected on a white and colored screen, then captured by a camera, respectively. Next, each ramp patch on colored screen is related with that on white screen which have the same digital values obtained by the camera. This procedure provides quantity of color shift between on white and colored screen. After that, a color correction matrix is obtained by regression method using the relationship above. In our experimental results, modified images using proposed method on colored screen have better performance than previous methods focused on objective and subjective evaluation.