JPEG Files

The Joint Photographic Experts Group (JPEG) has defined a family of compression techniques for working with photographic images. The most common of these use lossy compression, which means useless data is discarded during the compression process. JPEG compression was designed to discard data that the human eye cannot easily see. The encoded image will therefore not be an exact bit-for-bit representation of the original image, but it will achieve a high degree of compression with little or no perceived degradation in quality. JPEG compression works very well for photographic or scanned images, but it fares poorly on CAD drawings or "comic book" style pictures where there are large areas of the same color.

While the evolving JPEG standard defines various data compression schemes, it does not define a traditional file format for storing a JPEG-compressed image. What most people call JPEG files are, strictly speaking, JFIF files (JFIF stands for JPEG File Interchange Format). JFIF is a well-defined file format designed for storing images containing an encoded JPEG data stream. Because nearly everybody refers to JFIF files as JPEG files, we'll avoid potential confusion and do likewise in Fastgraph.

The most common JPEG files use a subset of the JPEG standard called baseline JPEG. A baseline JPEG file contains a single image compressed with the baseline discrete cosine transformation (DCT) and Huffman encoding; it must also use an 8-bit sample precision. The image may be 24-bit color (three components per pixel) or grayscale (one component per pixel).