Analysis of DVR Function in CCTV Surveillance System

The video recorder is an indispensable device in the closed-circuit television monitoring system. It can record the scene of the surveillance scene in real time and can be recorded in real time, and can be easily retrieved and checked afterwards to provide important clues and evidence for case detection. With the development of digital compression technology and large-scale integrated circuit technology, digital video recorders (DVRs) have developed rapidly in the television monitoring industry. There are not many domestic manufacturers that are capable of developing or producing DVRs on their own, and each DVR The structure and implementation are not the same.

The rapid development of digital hard disk recorders is mainly due to its own simple mechanical structure, using high-precision large-capacity hard drive as a recording device, therefore, as long as the image acquisition card is inserted in the computer expansion slot, coupled with the corresponding j monitoring System software and application software implement all the functions of a traditional tape recorder. In particular, the embedded HDD recorder that appeared later has a more compact structure and more stable performance, and it has almost become the terminator of the traditional time-lapse video recorder.

Realization of DVR

There are several ways to implement a hard disk recorder. From the system structure point of view, there is a PC plug-in card or embedded integrated model; from the core chip used, some are based on digital signal processors (DSP), and some are based on application-specific integrated circuits (ASIC), which based on DSP The structure is divided into different series, which are different due to the use of DSPs from different manufacturers; and from the technology (video compression format) of processing video for DVRs, it is based on Wavelet, M-JPEG, MPEG-1, MPEG-2. MPEG-4, H.263, H.264 and many other video compression formats. In addition, whether it is a PC card type or an all-in-one model, even if the chips they use are the same, the interface and function of the application software are not the same.

The earliest DVR was a PC card type. The video capture card mainly included video signal acquisition, digital video compression processing, and video buffering. Among them, the digital video compression processing chip had many different types (general DSP or dedicated ASIC). . With the continuous upgrading of core chips such as CPUs and memory, the frequency and integrated processing capabilities of computers have been continuously improved. As a result, a multi-card multi-channel hard disk recorder has emerged on the basis of single-card hard disk video recorders. A plurality of single-channel video and audio capture cards supporting parallel processing are simultaneously inserted in the expansion slot to achieve real-time acquisition of multiple channels of video and audio signals simultaneously. Since each card only corresponds to one signal, the number of acquisition cards can be flexibly configured according to the number of video signals. However, when multiple cards are inserted in the PC, the occupied PC resources increase accordingly, such as CPU and memory resources, the number of expansion slots on the motherboard, and the power supply of the motherboard. Therefore, when the number of camera sources (that is, the number of acquisition cards) is large, such hard disk recorders must use industrial computers with multi-slot industrial control boards, and be equipped with high-power power supplies, and have higher requirements on the frequency of the CPU.

In order to solve the resource occupation problem of multi-card applications, shortly after the single card single-channel hard disk recorder came out, some manufacturers introduced multi-channel video and audio capture cards that integrate two or even four video processing chips (DSP or ASIC) on one card. Therefore, real-time acquisition and compression of two-way signals or four-way signals can be realized at the same time. This structure is actually a video processing chip uniquely associated with each video signal, but they share a PCI-PCI bridge chip, thus occupying only one PC slot, plus the collection of video and audio signals is compressed by the card. Implemented by hardware, it effectively reduces the occupation of PC resources by HD video recording.

There is also a PC-based single-card multi-channel hard disk video recorder which is different from the above implementation principle: a video processing chip on a card needs to process multiple input signals, and thus it is necessary to collect multiple video signals in a time-division alternating manner. Record in M-JPEG compression format.

Quasi-embedded hard disk recorder based on PC structure

Although the compression efficiency of M-JPEG is not as high as the compression formats such as MPEG-1, MPEG-4, and H.264 based on multi-frame predictive coding, the frames of consecutive frames lose correlation when multiple channels of video are collected in single-channel rotation. Sex (it is not the picture taken by the same camera at all), so the use of inter-frame prediction based video compression algorithm loses its meaning and can only use intra-frame compression algorithms.

Therefore, this type of hard disk video recorder independently performs JPEG compression processing on each frame of the captured image, and then forms an independent M-JPEG file corresponding to each frame of input for each frame. This method can obviously facilitate multi-channel acquisition. For example, it can easily accommodate up to 16 channels of video input without regard to the continuity of the video frame, but only for video at a rate of 25 frames per second. For a video processing chip for signal acquisition, no matter how many video signals are switched to their input in turn, their “total resources” of 25 frames/s cannot be changed, so for this type of hard disk video recorder, each The maximum average frame rate of the road picture is only 25/16=1.56 frames/s (ideal value).

The improved products of the above structure increase the number of channels for video capture (eg, 4 acquisition channels on one card), so that multiple video input signals can be collected in parallel on each acquisition channel, which is equivalent to an increase in number. Total number of "resources" displayed and recorded (multiple rotation plus multi-channel acquisition). For example, a manufacturer uses two 8-channel capture cards to achieve 16-channel signal acquisition, so that the "total resources" of the DVR reaches 160 frames/s.

It should be noted that since the M-JPEG compression algorithm lacks inter-frame compression, the overall video compression ratio will be small, which will increase the amount of image storage (which of course will increase the hard disk overhead). For example, when obtaining the definition equivalent to the MPEG-1 image quality, the number of bytes per frame of the M-JPEG image is about 6K to 20KB, which is about 3 to 10 times that of the MPEG-1 image. In addition, DVR products using the M-JPEG algorithm are difficult to achieve simultaneous recording of multiple sound signals because the JPEG standard itself does not describe the sound compression method. Especially when the phenomenon of video dropout occurs due to multi-path rotation, how to synchronize sound is a problem that needs to be considered. In addition, because M-JPEG does not form a unified standard, but only a principle description or syntax specification of the compression method, so the actual M-JPEG standard is set and compiled by each DVR manufacturer. Each manufacturer The M-JPEG standard is not universal. This means that video files recorded by a brand's DVR are generally not callable in other DVR systems based on M-JPEG compression, nor can they be called by universal media player software such as Microsoft Media Player (Media Player). This limits the networking applications of multiple monitoring systems of different brands.

The DVR based on the PC card introduced in the previous section does not deviate from the PC architecture: the appearance of the PC, the architecture of the PC, the operating system of the PC, the interface of the PC,... Therefore, it can be considered as an extended application of PC. As long as the user exits the hard disk recording application (or puts the application running in the background), this DVR is a standard PC, and the user can easily perform the MS Office environment. Document editing, report statistics and other operations, .... However, because of this, this type of hard disk video recorder is very easy to be attacked by viruses and causes system crashes; it may also cause system downtime due to hardware compatibility issues or due to certain bugs in system software; It may even be due to the system administrator's own problems (for example, the operator puts the video program running in the background while playing games in the foreground) and the video recording system cannot be used because of some mis-setting and misoperation.

In order to break away from the PC system, some merchants have introduced a quasi-embedded hard disk recorder. But in essence, this DVR is not really out of the PC system, because it still uses the hardware structure of the PC: In addition to the CPU and other peripheral devices on the motherboard, it also integrates a graphics card, sound card, network card, but also for plugging The PCI expansion slot of the video capture card uses a high-power switching power supply with a fan as the power supply section. However, in order to make it appear in the monitoring market in the form of professional equipment, the machine has adopted an integrated design. Compared with the aforementioned PC-based DVR, this kind of quasi-embedded DVR fully utilizes the hardware resources of the PC and effectively considers the overall spatial layout of the machine. The structure is more compact and the volume is also effectively reduced, and is specifically used for monitoring. Hard disk recording in the system. original