This paper describes some methods on optimizing parallel processing and data structure for DCT
quantization
and Huffman coding of JPEG algorithm with parallel instruction of 32 bits fixed-point processors
in order to develop its parallel processing capability effectively.The result of simulation indicates that the methods obviously speed up parallel implementation of JPEG Algorithm
being applicable to the real-time image compression.