Variable length coding/decoding method for motion vectors

Abstract

In connection with compression-coding of video signals on the basis of inter-frame correlation, a single reference table is used for variable length encoding of inter-frame motion vectors established on the basis of various motion vector value ranges and degrees of accuracy. A reference table is provided for variable length encoding motion vectors based on a particular value range and degree of accuracy. In order to use the same table for motion vectors based on a larger value range than that for which the table was designed, the value of a motion vector to be encoded is divided to form a quotient and a remainder. An addition bit code is formed on the basis of the remainder and is appended to a variable length code which corresponds in the reference table to the quotient so that a variable length code value is formed for the motion vector based on the larger range. As to motion vectors based on a finer degree of accuracy than was provided for in the reference table, the motion vector value is multiplied by an appropriate factor and the resulting product is used to obtain a corresponding variable length code value from the reference table. Decoding is performed using the same table and by means of decoding operations that are the inverse of the encoding operations.