What Is Image Coding?

Image coding is the process of representing visual data in a compact, machine-readable form that reduces storage requirements and transmission bandwidth while preserving perceptual quality to a degree appropriate for the intended application. It encompasses the mathematical transforms, quantization schemes, entropy coding methods, and file format standards that convert a raw pixel array into a compressed bitstream. The field sits at the intersection of signal processing, information theory, and human visual perception, drawing on each to make principled choices about what data can be discarded without degrading the viewer's experience.

Image coding draws on transform theory established in the 1960s and 1970s, particularly the discrete cosine transform (DCT) and, later, the discrete wavelet transform (DWT). These transforms convert spatial pixel values into frequency coefficients that concentrate most of the signal energy in a small number of terms, making subsequent quantization and entropy coding far more efficient than operating directly on the pixel domain.

Still Image Compression Standards

The two most widely deployed standards for still image compression are JPEG and JPEG 2000, both developed through the Joint Photographic Experts Group working under ISO and ITU-T. Classic JPEG, standardized as ITU-T T.81 and ISO/IEC 10918-1, partitions the image into 8x8 pixel blocks, applies the DCT to each block, quantizes the resulting coefficients, and encodes them with Huffman or arithmetic coding. The quantization step table controls the trade-off between bit rate and fidelity, and compression ratios of 10:1 are routinely achieved on photographic content with minimal visible degradation. The ITU-T T.81 specification, hosted by the W3C, defines the normative encoding and decoding procedures for the JPEG baseline process.

JPEG 2000, standardized in ISO/IEC 15444, replaces the block-DCT with a discrete wavelet transform applied to the entire image, enabling scalable decoding at multiple resolutions and bit rates from a single compressed file. This scalability makes JPEG 2000 the preferred format in digital cinema (DCI), medical imaging archives, and remote sensing applications where a receiver may need only a reduced-resolution preview before requesting the full bitstream.

Video Coding and Transcoding

Video coding extends still-image compression by exploiting temporal redundancy across frames. Predictive coding estimates each frame from one or more previously decoded reference frames using motion compensation, encoding only the residual difference rather than the full pixel values. The MPEG family of standards, including MPEG-1, MPEG-2, MPEG-4, and the ITU-T H.264/AVC and H.265/HEVC standards, defines the syntax and decoding procedures for compressed video streams. A tutorial on image and video coding standards in IEEE Xplore surveys the progression from H.261 through H.265 and places each standard in the context of the applications that drove its design requirements.

Transcoding refers to the process of decoding a compressed bitstream and re-encoding it in a different format, at a different bit rate, or for a different target device. Transcoding is necessary when content encoded for broadcast distribution must be adapted for low-bandwidth mobile delivery or when legacy MPEG-2 material is being migrated to H.264 archives. The operation is computationally intensive because it requires a full decode-encode cycle and introduces additional generation loss if not managed carefully.

Applications

Image coding has applications in a wide range of fields, including:

• Digital broadcast and streaming media, where compression determines achievable video quality at a given channel bandwidth
• Medical imaging archives, using lossless or near-lossless compression to store radiology images in PACS systems
• Satellite and aerial remote sensing, transmitting large multispectral datasets over constrained downlinks
• Digital cinema and post-production, applying the JPEG 2000-based DCI standard to distribute digital cinema packages
• Secure communications, where steganographic techniques embed hidden data within image coding structures