System and method for scalable coding of sparse data sets

Abstract

A data encoding system and method successively generates compressed data on a bit plane by bit plane basis, starting with the bit position of the most significant non-zero bit for the node in the data array having the largest absolute value, and then encoding the data in the array for progressively less significant bits. All the nodes in the data array are represented initially by blocks of nodes on a block list, and later in the processing by nodes on two node lists. Whenever a block contains a node whose most significant bit is on the bit plane currently being processed, the block will be subdivided recursively until all the nodes in the block whose most significant bit in on the current bit plane are placed in a node list. Data bits representing an m.sup.th least significant bit of the block and node values are written to the compressed data file first, where m is the minimum number of bits required to represent the node having the largest absolute value in the entire data array being encoded. Data bits for successively less significant bit planes are written to the compressed data file until a bit plane stop point is reached. The bit plane stop point may be predefined, user selected, or procedurally selected (e.g., in accordance with available bandwidth for transmitting compressed image data).