Intraframe

In video compression, an intraframe, or I-frame, is a frame encoded without reference to any other frames. It is an intra-coded frame that contains all the information necessary to reconstruct that single image, using spatial redundancy within the frame rather than temporal information from neighboring frames. Intraframe coding typically relies on transform coding, quantization, and entropy coding to compress the image data.

I-frames serve as key reference points for decoding and for subsequent interframes, and they enable random access and frame-accurate seeking. Because they do not reuse temporal information from other frames, I-frames usually require more data to achieve the same quality as interframes. Video encoders insert I-frames at regular intervals, determined by the GOP (group of pictures) length, to balance quality, latency, and error resilience. In many codecs, including MPEG-1, MPEG-2, MPEG-4, and modern codecs like H.264/AVC and HEVC, I-frames are fundamental building blocks, with P-frames and B-frames providing temporal predictions to reduce overall bitrate between I-frames.

The concept of intraframe coding can also apply to still-image compression, where each image is encoded independently without reference to other images. In summary, an intraframe is a self-contained frame encoded independently, providing a reliable decoding reference while typically incurring higher data size than interframes for equivalent visual quality.