How JPEG DCT compression works: tiles, cosines, and discarded information

JPEG is a family of image compression methods, not a single compression algorithm. The original JPEG standard includes several approaches, including lossless predictive coding and lossy compression based on the discrete cosine transform (DCT). Later standards introduced additional methods. By far the most widely used JPEG methods are based on the DCT. The simplest and most familiar is called baseline JPEG. That’s the method we’ll examine here, and we’ll call it JPEG, for simplicity. JPEG compression takes advantage of two characteristics of photographic images: neighboring pixels are often similar, and the human visual system is less sensitive to some kinds of image detail than others. JPEG uses mathematics to represent image information in a form that makes it relatively easy to discard the less important parts. The heart of conventional JPEG compression is the discrete cosine transform, or DCT, which converts collections of pixel values into collections of spatial-frequency…

Commenting on Bubbles has been removed. Why? If you want to talk about this post, write to the author, use the comments on their blog if it has them, or reply with a "RE:" post on your own blog. Suggestions? feedback@bubbles.town