Why can videos be compressed more than images?

The primary reason that videos can be compressed more than images is due to the inherent differences in their data structures and the types of redundancy they contain. Here are the key factors that contribute to this difference:

1. Temporal Redundancy: Videos consist of a sequence of frames that often have very little change from one frame to the next. This means that a lot of information from one frame is repeated in subsequent frames. Video compression algorithms, like those used in formats such as H.264 or VP9, exploit this temporal redundancy by only storing changes between frames instead of storing each frame independently. Techniques like motion estimation and compensation are used to predict changes between frames, which significantly reduces the amount of data needed.

2. Spatial Redundancy: Both images and videos have spatial redundancy, which refers to the repetition of similar patterns or colors in different areas of a single frame or image. Techniques such as JPEG compression for images attempt to reduce this redundancy by using methods like the Discrete Cosine Transform (DCT) to focus on significant visual information. Video compression uses similar spatial compression techniques within each frame.

3. Psycho-Visual Redundancy: Videos can take advantage of psycho-visual redundancy, which refers to the limitations of human vision. Compression algorithms remove details that are less likely to be noticed by human viewers, such as subtle differences in colors or less perceptible details in fast-moving objects. This allows for more aggressive compression.

4. Inter-Frame Compression: Videos commonly use inter-frame compression, meaning that only key frames (I-frames) are fully stored, while other frames (P-frames and B-frames) are stored in terms of their differences from these key frames. This is a stark contrast to how images are compressed independently without relying on temporal data.

5. Higher Tolerance for Degradation: Videos often have more tolerance for minor quality degradation than still images, especially when viewed in motion. Compression techniques take advantage of this by applying more aggressive data reduction strategies that would be more noticeable and unacceptable in a high-quality standalone image.

Overall, these combined factors allow videos to be compressed much more efficiently than images, yielding significant reductions in file size while maintaining an acceptable level of visual quality.