Additive Color Mixing Model Based on Human Color Vision for Bayer-type Pixel Structures

This paper describes an analytical model of additive color mixing on display images in consideration of both Bayer type pixel structures and human color vision. The proposed model treats additive color mixing as the following phenomenon; color signals on a display unit attenuate because of the pixel structure and human color vision, and overlap with each other. First, an outline of the proposal is shown by using both the characteristics of human vision and the RGB signals from display images. For instance, this paper takes the case where red and green are mixed and look yellow. Secondly, the properties of RGB signals derived from a Bayer-type pixel structure, which is widely used in consumer products, are expressed as a frequency spectrum by the Fourier transform. Thirdly, the human characteristics of color vision are formulated as a Gaussian low pass filter. Finally, the additive color mixing process and secondary light colors are qualitatively discussed by using the analytical results of this proposal. As a result, the effectiveness and validity of this proposal for the optimal design of RGB signal strengths and pixel structures of display units are verified