Paper Title
Reduction Of Memory Errors For Images In Jpeg 2000
Abstract
JPEG 2000 is a widely used image coding standard for applications like Digital photography, video transmission
and medical imagery. In the process of saving the images, since it processes one entire frame at a time, it consumes large
power. During voltage scaling SRAM memory errors occurs in JPEG implementation. To reduce these memory errors,
Contourlet transform technique is implemented in this project. This technique provides a flexible multi resolution, local and
directional expansion for 2D images. These images are then transformed and sent through the Embedded Block Coding with
Optimal Truncation (EBCOT) where compressed images are obtained. By Inversing this transform technique, decompressed
image is obtained. Since this technique covers all the edges and smoothens the contours, we can reduce the loss of
information/data thereby reduce the memory errors. This transform is advantageous since it does not require additional
memory and circuit design is simple and more importantly reduces the power consumption and thus improves the image
quality