Source-linked AI summary
Jpeg Image Compression Using Discrete Cosine Transform - A Survey
A. M. Raid, W. M. Khedr, M. A. El-dosuky, Wesam Ahmed
TL;DR
Image compression is needed to reduce data for efficient image transmission and storage. This survey reviews lossy DCT-based JPEG compression for full-color still images, covering its components and process; it describes color conversion, block DCT, quantization, zigzag sequencing, and entropy coding.
Problem
Image compression addresses the need to reduce image data for efficient transmission and storage while retaining a good reconstruction.
Method
The paper surveys lossy JPEG compression using DCT, covering color specification, block processing, quantization, zigzag sequencing, entropy coding, and decompression.
Results
The survey describes JPEG as a widely used standard for full-color still-frame applications and explains its DCT-based baseline coding process.
Takeaways & Limitations
JPEG combines DCT, quantization, and entropy encoding to provide lossy image compression for digital image storage and transmission.
Abstract
from arXiv · showhide
Due to the increasing requirements for transmission of images in computer, mobile environments, the research in the field of image compression has increased significantly. Image compression plays a crucial role in digital image processing, it is also very important for efficient transmission and storage of images. When we compute the number of bits per image resulting from typical sampling rates and quantization methods, we find that Image compression is needed. Therefore development of efficient techniques for image compression has become necessary .This paper is a survey for lossy image compression using Discrete Cosine Transform, it covers JPEG compression algorithm which is used for full-colour still image applications and describes all the components of it.
1. INTRODUCTION
Image compression reduces data for efficient image transmission and storage by exploiting spatial, spectral, and psycho-visual redundancies. The survey introduces transform, quantization, and entropy-coding components and evaluates compression using compression ratio and reconstructed-image quality.
- Compression motivation: Image compression reduces redundancy so images require fewer samples for storage or transmission while preserving a good reconstruction.The paper identifies spatial, spectral, and psycho-visual redundancy as compression targets.
- Compression categories: Lossy compression sacrifices finer image details to save bandwidth or storage space, whereas lossless compression preserves the original decoded values.The paper distinguishes lossy systems from lossless systems based on whether image information is discarded.
- Compression components: Transform coding maps image data into coefficients that are then quantized and coded; the survey identifies DCT as an early transform-coding approach.Typical lossy systems also include a source encoder, quantizer, and entropy encoder.
- Performance criteria: Compression ratio compares original and compressed image sizes, while MSE and PSNR measure distortion and reconstructed-image quality.The paper presents CR = n1/n2 and identifies PSNR as a widely used image-quality measure.
2. DCT TRANSFORMATION
The survey presents DCT as a widely used image-compression transform and explains that the paper next covers JPEG compression and decompression. DCT represents image data in a frequency domain where lower-frequency energy can support compact coding.
- DCT transformation: DCT is an orthogonal transform that maps image-space data into frequency components using a fixed set of basis functions.The paper describes DCT as popular in image and signal-processing applications.
- DCT transformation: DCT packs image energy into lower frequencies and can reduce visible blocking artefacts between sub-images.These properties are presented as advantages motivating DCT use in JPEG.
- JPEG transition: The paper explains the basics of JPEG compression and decompression after introducing DCT transformation.The following discussion covers JPEG rather than developing a separate transform method.
3. JPEG COMPRESSION
JPEG baseline coding applies DCT-based lossy compression to color images through color conversion, block transformation, quantization, zigzag ordering, and entropy coding. The survey describes corresponding decompression operations that reconstruct 8 × 8 blocks.
- JPEG standard: JPEG is an ISO/ITU-T standard targeted at full-color still-frame applications, with baseline, lossless, progressive, and hierarchical modes.The DCT-based baseline mode is described as suitable for most compression applications.
- JPEG process steps: JPEG color-image processing converts RGB to YCbCr, divides components into 8 × 8 blocks, shifts values to [-128,127], applies DCT, quantizes, and entropy-encodes coefficients.The color model separates luminance Y from chrominance Cb and Cr for processing and transmission.
- Quantization: Quantization divides each DCT coefficient by a quantization-matrix value and rounds the result, reducing many less-important high-frequency coefficients to zero.JPEG retains lower frequencies more strongly because the paper states that human vision is more sensitive to them.
- Zigzag sequencing: Zigzag sequencing places lower-frequency coefficients before higher-frequency coefficients, producing similar data bytes that support efficient entropy encoding.Higher-frequency coefficients are typically zero or nearly zero after quantization.
- Entropy coding: Huffman coding compactly encodes quantized DCT coefficients using symbol frequencies and a Huffman tree, and it is used in JPEG’s baseline sequential codec.The paper states that Huffman coding achieves optimal average length when the prefix condition is met.
- Decompression: Decompression restores Huffman tables, decodes tokens, reconstructs coefficient blocks with zeros, reverses zigzag order, and applies IDCT to recreate 8 × 8 blocks.The decompression phase reverses the compression phase in the opposite order.
4. CONCLUSION & FUTURE WORK
The survey focuses on fast, efficient lossy JPEG compression and decompression using the Discrete Cosine Transform, while introducing core compression principles and JPEG processing steps. Future work will compare DCT with the Discrete Wavelet Transform.
- The survey examines fast, efficient lossy JPEG coding and decoding using the Discrete Cosine Transform.
- It introduces digital image compression principles and various image compression methodologies.
- JPEG processing is described through DCT, quantization, and entropy encoding.
- Future work will compare Discrete Cosine Transform and Discrete Wavelet Transform techniques.