An experimental comparision of vector quantization techniques for Thai text image

This paper presents the experimental comparison of vector quantization (VQ) methods for Thai text image. The compared methods include of Linde-Buzo-Gray (LBG), Enhanced LBG (ELBG) and Genetic algorithm (GA). The performance of VQ over gray image is measured with peak signal-to-noise ratio (PSNR) and processing time. The studied parameter are including of codebook number, cluster size, training image type, number of training image and noise. In addition, the results are shown in the verities aspects which are conducted to the appropriate approach in compression of Thai text image.