Image data compression by the discrete cosine transform

The application of the Discrete Cosine Transform to image data compression is presented. In particular a fast algorithm is described for an efficient implementation of the transform. Both zonal and threshold sampling methods are considered and their performance is evaluated. Also a comparison of the performance of the Discrete Cosine Transform and the Fast Fourier Transform is outlined. Quantitative and qualitative results are presented for evaluating the various alternatives for the application of the Disscrete Cosine Transform. The results have shown that a convenient choice is the threshold sampling method applied to transformed blocks of size ranging from 8 × 8 to 32 × 32.