Sub-optimal truncation in JPEG2000 imagery for rate control and high compression throughput

This paper presents a method for implementing a post compression truncation algorithm in JPEG2000 used for rate reduction and control based entirely on coding pass rates from the Tier 1 coder. This sub-optimal truncation (SOT) method results in higher quality imagery than quantization of coefficients prior to encoding while allowing for rate control that mere quantization does not. SOT requires the coding pass rates to be recorded during the Tier 1 coding process and does not require the distortion calculation that traditional optimal truncation (OT) uses. SOT also implements a straight-forward truncation process that does not require an algorithm for optimization or for efficient implementation of coding pass truncation, reducing the computational requirement for a rate controller. Testing of SOT on standard imagery shows a compression gain within 0.1 dB of OT, and an computation reduction in overall JPEG2000 compression time of 5%, on average.