Experience HD: Principles and technologies for a clearer perspective

Traditional standard-definition surveillance is divided into three types: analog, digital, and network. The front end of analog surveillance is an analog camera, and the back end is a matrix. The front end of digital surveillance is also an analog camera, and the back end is a DVR. This is not the case with high-definition surveillance.

HD monitoring clear principle

For video surveillance, image clarity is undoubtedly the most critical feature. The clearer the image, the more obvious the details, the better the viewing experience, and the higher the accuracy of intelligent and other application services. Therefore, image clarity is the eternal pursuit of video surveillance.

The standard for measuring image clarity is resolution, which is measured in pixels. The larger the value, the clearer the image. The difference between high definition and standard definition is reflected here. The dividing line between the two is megapixels or 720p, and high definition is achieved when it reaches megapixels or 720p. Based on this standard, CIF and D1, which currently dominate the video surveillance market, are both standard definition.

Whether in terms of resolution, display effect or fluency, HD has more advantages than SD. In terms of resolution, the resolution of 720p is 9 times that of CIF, and the resolution of 1080i/1080p is 20 times that of CIF. Under the same display environment, HD will be much clearer. In terms of display effect, HD supports both large-screen display and 16:9 widescreen display, which can greatly enhance the user's viewing experience. In terms of fluency, HD supports higher frame rates. For example, 720p and 1080i/1080p can support 60 frames/second or 60 fields/second, and its image fluency is twice as high as that of SD. Therefore, HD surveillance will inevitably replace SD surveillance.

To achieve true high-definition monitoring, high-definition must be fully supported from the acquisition of video sources, encoding and compression of video signals, transmission of video signals, video browsing, playback of video files, etc. For customers, high-definition is meaningful only when it includes all links such as front-end, platform, storage, browsing, and display.

There are two transmission modes for the uncompressed HD video signals after acquisition: analog and digital. Analog transmission generally uses YPbPr component transmission, and one HD video signal requires three coaxial cables to be transmitted simultaneously. Digital transmission generally uses DVI, HDMI or HD-SDI transmission. The transmission distance of DVI or HDMI is only a few meters, which is not suitable for monitoring transmission. Although HD-SDI can be transmitted about 100 meters, it has high requirements for coaxial cables and the price of cables is also very expensive. However, if the HD video is efficiently compressed at the front end and then transmitted through the IP network, its transmission cost is equivalent to the transmission cost of standard-definition monitoring network.

At present, the video signal processing of DVR is divided into two parts: video acquisition and video encoding. In the video signal acquisition part, most DVRs currently use analog video input, and all use CVBS (composite video signal) signal input with BNC interface. The maximum resolution of the composite video signal is D1, and the composite video signal is acquired and quantized according to the D1 resolution inside the DVR. At present, DVRs do not have YPbPr analog component interface, or digital video interface such as DVI, HDMI, HD-SDI, etc. In the encoding part, DVR can encode each acquired video signal into resolutions such as CIF, 2CIF, DCIF, D1, etc., but the maximum encoding resolution can only be D1, and it does not have the encoding capability of 720p, 1080i, or 1080p. Therefore, the current DVR cannot achieve high-definition monitoring.

It can be seen that to achieve high-definition monitoring, high definition must be considered from the perspective of the entire monitoring system, and networking is the basis for the application of high-definition monitoring systems.

HD network camera for a clearer view

We know that traditional standard-definition surveillance is divided into three types: analog, digital, and network. The front end of analog surveillance is an analog camera, and the back end is a matrix. The front end of digital surveillance is also an analog camera, and the back end is a DVR. There are two types of front ends for network surveillance, one is an analog camera + video encoder, and the other is a network camera, and the back end is a platform. In other words, in the era of standard-definition surveillance, there are many types of front ends, such as analog cameras, network cameras, and video encoders + analog cameras.

So why, in the era of high-definition surveillance, the front end must be a high-definition network camera? The reasons are mainly in two aspects.

First, high-definition surveillance must be networked and must be based on video compression processing and transmitted through an IP network, because only in this way can the transmission cost of high-definition video be controlled.

There are two transmission methods for uncompressed HD video signals after acquisition: analog and digital. Analog transmission generally uses YPbPr component transmission, and one HD video signal requires three coaxial cables to be transmitted simultaneously. Digital transmission generally uses DVI, HDMI or HD-SDI transmission. The transmission distance of DVI or HDMI is only a few meters, which is not suitable for monitoring transmission. Although HD-SDI can be transmitted about 100 meters, it has high requirements for coaxial cables and the price of cables is also very expensive. Whether it is transmitted in analog or digital mode, the transmission cost of uncompressed HD video signals is significantly higher than that of previous analog standard-definition video signals. When video compression encoding is used and transmitted over an IP network, the difference in transmission cost between HD video and standard-definition video is very small.

Second, the efficiency and cost of directly using HD network cameras for front-end processing are more advantageous than using HD video encoders + HD cameras.

Currently, CMOS is the main type of high-definition video sensor. CMOS sensors directly output digital video signals. In the camera, the digital high-definition video signals are directly compressed and encoded through DSP or ASIC, and then transmitted via a network. This is more economical than the camera directly outputting high-definition signals, and has higher processing efficiency.

This is why all the high-definition surveillance front ends we can see on the market are high-definition network cameras.