LAUSR

Abstract Spectrum is the essence of color, which can retain the main information of color. Spectral color information is generally used in the liquid crystal displays (LCDs) color field. Color spectrum of LCDs is usually a very large file, but an appropriate compression method can significantly reduce the required storage space while minimizing the loss of information. In this paper, a color difference weighted principal component analysis (CDWPCA) method for LCDs spectral compression is proposed. First, we investigated the importance of wavelength to color information. Then, a weighted function considering wavelength change and human visual system is introduced into spectral compression. The performance of CDWPCA are tested by comparing the results of standard principal component analysis and other weighted function principal component analysis by compressing and reconstructing the test samples from the eigenvectors of different training samples. The results show that the proposed CDWPCA method can significantly improve the retention rate of color information in the compression process of principal component analysis.