Digital Video Compression Techniques

Digital video compression (or video encoding) is the process of cutting the size of video files and changing them to proper formats and codes for viewing. The prime aim of encoding is to reduce the amount of data required to record, store, and transfer videos. As a result, storage hardware, data transfer times, and necessary delivery bandwidth are reduced.

Digital video compression is a group of methods used to reduce and eliminate inconsistencies in video data. This results in the compressed video being much less in size than the uncompressed video. Video encoding helps to store the video in reduced file size or sends it more rapidly over a network. The effectiveness of digital video compression is proportional to the video bitrate for a fixed quality and frame rate (fps). If the encoding results in lower bandwidths, it is more efficient.

How Video Compression Works

After compressing or encoding a video file, the compressed video can either be the same quality as the uncompressed file or may have reduced quality. These are called lossless and lossy compression formats.

Lossy Compression Formats

This digital video compression format ensures the compressed file contains less data than the original. Colors and sounds that recur throughout the video may be excluded to chop out sections of the film that are deemed unnecessary. Sometimes, this results in a low file because data is lost, hence the term "lossy." Lossy compression causes the data to be reduced to a level at which original information can't be retrieved after decompression. The difference between the encoded video and the original is referred to as the artifacts.

Most online video files employ lossy compression to reduce file size while maintaining a high level of quality. If a video kept its exceptionally high-quality file size, it would not only take a long time to upload the video, but viewers with poor internet would have a difficult time viewing or downloading it.

Lossless Compression Formats

Lossless compression is another digital video compression format where the original and compressed versions are almost identical. It is a method that allows the original data to be rebuilt from encoded data. That is, a restricted number of approaches are accessible for data reduction, resulting in limited data compression. In the encoding process, no data is lost. Most times, lossless compression formats are not as beneficial as lossy compression since records often wind up having the same size as before compression. GIFs are an example of lossless compression. However, because of its limitations, it is unsuitable for surveillance footage.

Adopting lossless compression may appear to be unnecessary, considering that the major purpose of compression is to reduce file size. Nonetheless, lossless compression produces excellent video quality if the file size is not a concern.

Digital Video Compression Techniques

Video data can be displayed as a sequence of static frames or fields for interlaced video. The picture sequence will undoubtedly include both temporal and spatial redundancies, which digital video compression methods can exploit. Many video encoding techniques employ temporal compression, which is centered on redundancy between video frames, and spatial compression, which is focused on redundancy inside a single frame or field.

Spatial Compression

Spatial compression methods are based on static image compression. The transform approach is the most widely used method and is used by various organizations. This method divides the picture into chunks and applies the transform to each block individually. The transform's output is resized and quantized. Then, a lossless entropy coder compresses the quantized data, generating the resulting bit stream. The most commonly used transform technique is the Discrete Cosine Transform (DCT) or its variations. Other spatial compression methods include fractal compression, vector coding, wavelet transform, etc.

Temporal Compression

Temporal compression is another effective digital video compression technique. Temporal compression operates by analyzing distinct video frames to one another. If there are no moving sections in the video, the program can execute a brief instruction that duplicates that section of the previous frame bit-for-bit into the following one. If the algorithm notices some pixels are changed, it employs a prediction approach in relation to the previous frame or frames. The program searches for a comparable region in the preceding frame or frames to every region in the current frame. If a comparable area is located, the difference is translated by the transform coder and removed from the current area. The reference for the current frame area can alternatively be computed as a weighted sum of comparable areas from previous and subsequent frames. The current frame must be deferred by a certain number of frame intervals when using successive frames.

Other commonly used video compression techniques include:

Joint Photographic Experts Group (JPEG) Compression

Paired with the MPEG compression technique, it is a basic digital video compression technique that works on a single (still) frame or picture. It is one standard of video encoding proposed by the International Organization for Standardization (ISO), the International Electrotechnical Commission (IEC), and the ITU (International Telecommunication Union). An organization named the Joint Photographic Experts Group (JPEG) was established in the middle of the 1980s, laying the groundwork for the JPEG standard. The group's first visible contribution was the 1991 publication of the first component of the JPEG standard, which it had set out to create as a standard for color photo compression; ever since, the JPEG group has worked on both the original JPEG standard and the JPEG 2000 standard.

Assuming we photographed an image with a guy on a motorcycle and trees in the background, JPEG techniques encode the image by removing elements that are not noticeable to the human eye, such as certain colors (since we are not as perceptive to them - brightness adjustments are more perceptible to humans than color variations) and some specifics in the more intricate areas of the image (changes in low-frequency lights are more noticeable than changes in lights of high frequencies).

Depending on the compression rate used, any data we delete may affect the quality of the resultant image. That is why JPEG is known as a lossy compression method, implying that some original picture information is lost and cannot be recovered.

Motion JPEG

Suppose we take a video of this identical situation with the cyclist pedaling. The camera takes a series of still pictures, which can be compressed using JPEG algorithms.

This digital video compression technique is called Motion JPEG. It was widely used in the early 1990s. It efficiently compresses video by removing spatial redundancy, but it ignores picture elements that are shared across frames, such as the backdrop, the tree, and so on. This shared material is called temporal redundancy, which can be deleted from video to improve compression rates.

Moving Picture Experts Group (MPEG)

This digital video compression technique, along with JPEG, is the basic digital compression technique proposed by ISO, IEC, and ITU. The genesis of MPEGs is the creation of the Motion Picture Experts Group (MPEG) in the late 1980s, which aimed to develop a system for the compression of motion pictures and sounds. Since then, it has generated standards for MPEG 1, MPEG-2, and MPEG-4, as well as standards unrelated to multimedia codings, such as MPEG-7 and MPEG-21.

Algorithms for inter-frame compression were developed to exclude temporal redundancy (inter-meaning between or among). MPEG is the most widely used of these. MPEG compression, like JPEG, is done in fundamental units called "macro-blocks."

Each frame is broken into slices, each of which comprises macro-blocks. Macro-blocks comprise pixels blocks, the picture's smallest modifiable part. In the compression process, macro-blocks are employed as a fundamental object.

MPEG frames can be compressed in three ways:

● Intra-coded (I-frames): Videos are encoded as isolated frames (stationary frames) that are unrelated to subsequent frames.
● Predictive-coded (P-frames): The algorithm encodes the videos using a prediction from a previous I-frame or P-frame, resulting in a higher compression ratio (smaller frame).
● Bi-directional-predictive-coded (B-frame): This process involves encoding a video by prediction, using a previous and a future frame from either I-frames or P-frames; it provides the most compression.

Other advanced MPEG formats are MPEG 2 designed specifically for TV broadcasting and high-quality archiving, and MPEG 4 used for texture mapping of 2D and 3D meshes, encoding time-varying streams, and making algorithms for temporal, spatial, and quality scalability, images, and videos, MPEG 7 for video surveillance and MPEG 21.

H.264

In the late 1990s, a new group called Video Coding Expert Groups (VCEG) created a video compression standard that would improve the efficiency of video encoding much higher than other existing techniques. This gave birth to H.264. It offers the most recent balance of code efficiency, cost, and computational complexity. The aim was to give enough versatility for the digital video compression technique to be used in many applications, including low and high data rate video, low and high-resolution video, and high and low latency needs.