Effective Channel-Independent Inverse Characterization Method for Display Device

摘要

This paper proposes a channel-independent inverse characterization process based on the GOG model for a display device. The CIEXYZ values for each of RGB channel can estimated from nine channel-independent interpolation TRCs (Tone Response Curves), and the result of this characterization method is better than that from conventional three channel-independent interpolation TRCs. However inverse characterization is impossible when using nine channel-independent interpolation TRCs, as the CIEXYZ values that correspond to each RGB values are inseparable directly. Accordingly, inverse characterization is usually implemented using the 3D-LUT (Look-Up Table) method. Yet, although the result of a 3D-LUT is accurate, creating the LUT requires a lot of memory space and considerable amount of measurements. Therefore, an accurate inverse characterization method is proposed based on the simple modeling of channel-dependent values and nine channel inverse processes based on the GOG model. CIEXYZ values are computed for three normalized luminance values using an inverse matrix. The channel-dependent values are subtracted from the normalized luminance values that are generated by an overlapped spectral distribution of the primary digital values based on modeling the channel-dependent values. Each of the three normalized luminance values is modified into the corresponding nine channel TRCs using a gamma correction of each channel and modifying the tone response curve. The digital values are estimated by the nine channel-independent interpolation TRCs based on an inverse GOG model using the parameters of forward characterization, considering weighting factors. Three digital values are determined for each RGB channel based on the maximum CIEXYZ values for each red, green, and blue channel. As such, the proposed method reduces the time complexity and number of measurements required for accuracy.