Understanding JPEG Compression

Why JPEG?

JPEG is the most widely used image format in the world. It makes photos smaller without losing too much quality. This is important because smaller files save storage space and load faster on the internet. Learning how JPEG works is also valuable in computer vision, multimedia, and AI systems, where image processing is a core task.

What I Did 

In this study, I explained how JPEG compression works step by step:

• Color Conversion: Images are converted from RGB to YCbCr, because the human eye is more sensitive to brightness (Y) than color (Cb, Cr).
• Block Division: The image is split into small 8×8 pixel blocks.
• Discrete Cosine Transform (DCT): Each block is transformed into frequency values.
• Quantization: Higher frequencies are reduced, because the eye notices them less.
• Zig-zag Ordering: The values are ordered from most important to least important.

• Huffman Coding: The data is encoded in a way that reduces size further without losing information.

I also studied the advantages and disadvantages of JPEG. For example, it is efficient and easy to use, but it is lossy, which means some detail is always removed.

Practical Tasks

This project was not only theory but also practice.

1. I worked with image filters (Gaussian, mean, median), normalization, and edge detection. These steps are important for preparing data before compression or analysis.
2. In the final report, I focused on JPEG compression itself, explained the pipeline in detail, and compared methods.
3. I also wrote a research article, where I reviewed literature and explained the math behind DCT, quantization, and coding.
4. Finally, I prepared an introduction that gave background information and explained the motivation for the study.

Through this project, I understood how mathematical concepts like transforms and probability connect with real-world needs like reducing file size. I also saw how human vision plays a role, because JPEG is designed around what the human eye can and cannot see easily.