Variable length-adaptive image data compression method and apparatus

A method for compressing variable length input image data comprising the steps of:

• (a) dividing the input image data into a predetermined number of blocks and calculating a DC coefficient and an AC coefficient for each block;
• (b) calculating an activity level for each block, to represent the complexity of each block, based on said AC coefficient for each block;
• (c) classifying the blocks into a predetermined number of classes, each having the same number of blocks, based on the calculate, d activity level for each of the blocks;
• (d) quantizing and encoding said DC coefficient for each block into DC-encoded data formed of a number of DC bits for each block;
• (e) estimating a quantity of AC bits for encoding the AC coefficient for each block based on the number of DC bits for each block, based on a quantity of overhead bits representing overhead-encoded data derived from encoding overhead information and based on a total number of bits allocated to each block in each class, and selecting an optimal quantizer for generating the least error based on the estimated quantity of AC bits;
• said estimating step including,
• (e-1) inputting the quantity of DC bits used for coding the DC coefficient of each block, the quantity of overhead bits used for coding overhead information, and the total number of bits allocated to the block;
• (e-2) allocating an allocated quantity of AC bits for encoding the AC coefficient in each block by subtracting the quantity of DC bits and the quantity of overhead bits from the total number of bits allocated to the block;
• (e-3) quantizing the AC coefficient for each block by a quantizer scaled by a largest value CRO) in a prepared quantizer range scaled factors (Rm) set and carrying out a variable length coding of the quantized AC coefficient by using a truncated Huffman code table;
• (e-4) calculating a calculated quantity of bits used for the variable length coding of the step (e-3) and comparing the calculated quantity and the allocated quantity;
• (e-5) carrying out the steps (e-3) and (e-4) iteratively, while varying the scale factor in the quantizer range scaled factor set, until said calculated quantity most closely approximates said allocated quantity without said calculated quantity exceeding said allocated quantity;
• (f) quantizing and encoding said AC coefficient for each block using said optimal quantizer to form AC-encoded data;
• and
• (g) combining said DC-encoded data, said AC-encoded dam and overhead-encoded data to form a compressed data output.