TV cameras with hybrid compression technology.
IP cameras using a hybrid compression algorithm that provides intra-frame and inter-frame compression of video footage are increasingly conquering the network video surveillance market.
Hybrid (stream) compression algorithms include the MPEGX (MPEG1, MPEG2, MPEG4, etc.) and H.26X (H.261, H.263, H.264, etc.) group algorithms (codecs). The algorithms are very similar in their fundamental principles, as they use the discrete cosine transform (DCT) or its modifications to eliminate spatial redundancy. In fact, the algorithms have different names due to patenting in two different standardization organizations (expert groups) – ITU-T (International Telecommunication Union, H.26X group standards) and ISO/IEC (International Organization for Standardization/International Electrotechnical Commission, MPEGX group standards). Some of them, such as MPEG4 part 10 and H.264, are technically completely identical.
In this article I will not give a detailed description of each of the algorithms and its implementation. This information has been published many times on the pages of various publications.
The fundamental fact is the higher compression ratio of streaming video by hybrid codecs compared to only intra-frame compression using the MJPEG algorithm. Unlike intra-frame compression of MJPEG, hybrid algorithms are based on the ability to predict the movement of an object from frame to frame and, as a result, redistribute redundancy in space.
The table shows the average compression ratios typical of the most popular compression algorithms. As you can see, the efficiency of hybrid compression methods compared to MJPEG can reach 6 times.
Hybrid compression algorithms have long been used in digital recording devices — digital video recorders (DVR), which is primarily due to the need for efficient use of disk space for long-term recording. An important factor is the relative simplicity and low cost of implementing hybrid compression in DVR.
Here it is necessary to understand that the compression efficiency of a particular codec is directly related to the amount of calculations performed by the processor providing compression. The more efficient the codec, the more complex the calculations, and the more powerful the processor performing these calculations must be. For calculations according to the MPEG algorithm for several video signals, the resources of one DVR processor can be used.
However, for each IP camera, unlike a DVR as a separate image source, it is necessary to use a powerful and expensive data compression processor. This is especially true for IP cameras with megapixel matrices, where the data volume is several times greater than the data volume of analog resolution matrices. It was the release of new and reduced cost processors that stimulated the production of IP cameras with hybrid compression technology. It should not be forgotten that decoding (analog representation), for example, an MPEG4 video stream, requires a more powerful computer (workstation) than decoding an MJPEG stream with similar characteristics (resolution/frame rate).
Thus, the main advantage of IP cameras with hybrid compression technology is the efficiency of using the transmission network bandwidth, since the volume of data transmitted by the camera is fundamentally lower than that of IP cameras with the MJPEG codec. Thus, with the beginning of the use of hybrid compression algorithms in IP cameras, a real opportunity has appeared to create large network video surveillance systems with hundreds of cameras.
Another important aspect of low network load is the greater flexibility of using IP cameras in various network environments, especially low-speed ones. We are talking about the ability to transmit a larger number of frames, i.e. providing more information about the object of observation. This means that on the data transmission channel, where an IP camera with the MJPEG codec provides transmission of individual image frames, an IP camera with a hybrid compression algorithm will allow receiving streaming video.
The most common hybrid compression algorithm for IP cameras is currently MPEG4. A large number of processors providing the necessary calculations and their relatively low cost allowed manufacturers to finally abandon MJPEG encoding as a profile codec. At the moment, MJPEG is used as an additional codec. The MJPEG stream is present in IP cameras as an alternative. The fact is that MPEG4 is a proprietary algorithm, i.e. there are many registered MPEG4 profiles from various manufacturers, and they all have minor differences, do not support any universal software or playback tools. This means that to work with the MPEG4 stream, a third-party software manufacturer must use a patented codec. In turn, MJPEG, being a frame-by-frame compression method with intra-frame compression using the JPEG algorithm, is widespread and open.
In fairness, it should be noted that MJPEG is currently used for image compression in megapixel IP cameras, where the volume of required calculations is so large that it requires the use of additional coprocessors, and, consequently, an increase in the cost of the camera.
Unfortunately, with the development of new, even more effective compression algorithms, the issue of technical implementation of compression and its cost is especially acute. Thus, the emergence of the H.264 standard (MPEG4, part 10), especially relevant for compression of megapixel resolution video streams, is accompanied by a certain shortage of powerful and inexpensive processors. As a result, a number of manufacturers are trying to implement this algorithm on the same hardware base that was used for earlier MPEG4 profiles. The result is the absence of a fundamental difference (bit rate, number of frames) between the two algorithms.
However, this is a matter of time and in the near future, the development of IP cameras will be marked by a transition to H.264 as the most effective hybrid algorithm for video stream compression for network video.