DATA PROCESSING METHOD OF DISCRETE COSINE TRANSFORM (DCT), DCT METHOD, AND DCT DATA PROCESSING CIRCUIT

摘要

PURPOSE: To provide a circuit which is small-sized and has a high processing capability by using specific formulas I and II to perform the data processing of DCT caused between the table of N×M data f(x, y) and a table of M×N coefficients (u, v). CONSTITUTION: In the formula I, K is a constant-power of 2, and x, y indicates coordinates of data in the data table, and u, v indicates coordinates of coefficients in the coefficient table. In the formula II, c(u), c(v) is 1/21/2 for u, v=0 and is one for u, v≠0. Each coefficient F(u, v) is successively determined even for a pair of coordinates u, v, and the absolute value and the sign of a coefficient P=p(x, u), p(y, v) are searched in the table for a first pair of x, y. When the absolute value of P is multiplied by f(x, y), the result of multiplication is introduced to an accumulator by addition or subtraction with the function of the sign. An operation is restarted for all values of the pair x, y, and the result given by the accumulator is extracted.