Indian Journal of Science and Technology

Abstract

Background/Objectives: This paper presents a color image compression method to increase the compression ratio without affecting the original scene by noise or distortion. Methods/Analysis: In this paper an attempt to minimize data loss without highly affect the compression ratio by employing low lossy rate quad-tree compression technique to increase the correlation between pixels that will enhance DCT results and also compress the data before DCT phase, using Quantization and entropy encoders such as run length encoding and shift encoding will further compress the image. Finding: For conducted six different visual information images, the Compression Ratio (CR) results of the proposed method on average to be around 1:29 of the size of the original image, more compression ratio can be achieved by increasing the compression levels, this high compression ratio is considered a best ratio compared with the achieved Peak Signal to Noise Ratio (PSNR) of the decompressed-compressed image. Application/Improvements: This compression system can be used in Storing or Transforming Color Images due to its high compression ratio while the noise and distortion still as minimum as possible.
Keywords: Color Image Compression, Discrete Cosine Transformation, Quad-Tree, Quantization, Run Length Encoding, Shift Encoding