2. Introduction to Image Compression
Definition :-
Image compression is the process of reducing the size of an
image file without significantly degrading its quality.
Why compress images?
I. To save storage space
II. To reduce bandwidth for faster transmission
III. To optimize web performance and mobile apps
IV. To reduce cost for cloud storage and data usage
3. History of Image Compression
1960s–1970s: Early Techniques :-
I. Simple methods like Run-Length Encoding (RLE) and
Predictive Coding.
II. Focus on reducing data redundancy in black-and-white
images.
1996: PNG Format :-
I. A lossless compression format.
II. Replaced GIF; supports transparency.
III. Uses the Deflate algorithm.
4. History of Image Compression
JPEG 2000
I. Based on Wavelet Transform (DWT).
II. Better quality, but less adopted due to complexity.
2010s: New Web Formats
I. WebP (Google) – efficient for web, supports both
lossless/lossy.
II. AVIF – modern open-source format with high efficiency.
2020s: AI-Based Compression
I. Use of deep learning for smarter, perceptual-based
compression.
5. Need for Image Compression
Faster transmission :- Compressed images load quicker
on web/mobile.
Images take a lot of space :- Uncompressed images like
BMP or TIFF can be very large.
Limited bandwidth environments :- Required in
streaming and mobile networks.
Archiving :- Large image datasets (e.g., medical imaging)
need efficient storage.
6. Types of Image Compression
Lossless Compression :-
1. No loss of data or image quality
2. Image can be fully reconstructed
3. Suitable for critical applications (e.g., medical imaging)
Lossy Compression :-
1. Some data is lost, can't be recovered
2. Much smaller file sizes
3. Acceptable for everyday images (e.g., web photos)
7. Lossless Compression Techniques
Run-Length Encoding (RLE) :-
1. Stores repeating values as (value, count)
2. Good for images with large areas of same color
Huffman Coding :-
1. Replaces common patterns with shorter codes
2. Used in PNG and JPEG formats
LZW (Lempel–Ziv–Welch) :-
1. Dictionary-based
2. Used in GIF, TIFF, and PNG
8. Lossy Compression Techniques
Color Space Transformation :-
1. Convert RGB to YCbCr
2. Reduces data for chrominance channels
Discrete Cosine Transform (DCT) :-
1. Converts image to frequency domain
2. High-frequency data (less visible) is discarded
Quantization :-
1. Approximate similar values
2. Main reason for data loss in JPEG
10. Applications
Web Development :- Fast-loading websites, SEO.
Medical Imaging :- Efficient transmission of large scan
files.
Satellite Imaging :- Store and send high-res photos from
space.
Mobile Devices :-Save space and battery life.
Multimedia :- Image storage in videos, animations.
11. Advantages & Disadvantages
Advantages :-
1. Saves disk space
2. Faster image transfer and loading
3. Lowers costs for cloud and web hosting
Disadvantages :-
1. Lossy methods reduce image quality
2. Some compression artifacts may appear (blurring, blocking)
3. Cannot fully recover original image from lossy formats
12. Future Scope
AI & Machine Learning Integration :-
Deep learning models (CNNs, GANs) will optimize image
compression based on content and human perception.
Real-Time Compression :-
Required for video conferencing, live streaming, and
augmented reality (AR).
Web & Cloud Optimization :-
Demand for faster website loading and lower bandwidth
usage is rising.
Secure & Privacy-Aware Compression :-
Emerging need to compress without compromising sensitive
data.
13. Image Compression Tools & Libraries
Online Tools :-
i. TinyPNG, Compressor.io, JPEG-Optimizer
Programming Libraries :-
i. Python: Pillow, OpenCV, PyTorch (for AI compression)
ii. JavaScript: Compressor.js, browser APIs
iii. C/C++: libjpeg, libpng
14. Conclusion
Image compression is essential for modern computing
Choice between lossless and lossy depends on use case
Understanding algorithms helps in selecting right format
Future trends are heading towards AI-powered
techniques
15. References
Sukhwinder Singh, Vinod Kumar, H.K.Verma, Adaptive
threshold based block classification in medical image
compression.
SalehaMasood, Muhammad Sharif, MussaratYasmin,
MudassarRaza and SajjadMohsin. Brain Image
Compression.
Ajit Singh, MeenakshiGahlawat, " Image Compression and
its Various Techniques.
16. References
S. Kumar, T. U. Paul, A. Raychoudhury, “Image
Compression using Approximate Matching and Run
Length”, International Journal of Advanced Computer
Science and Applications.
S. Kumar, T. U. Paul, A. Raychoudhury, “Image
Compression using Approximate Matching and Run
Length”, International Journal of Advanced Computer
Science and Applications, Vol. 2, No. 6, 2011.
