At present, the network camera video compression algorithms mainly include: M-JPEG/MPEG4/H. 264, while the domestic and H. 264 is the mainstream, basically has achieved the compression level of MainProfile, but each manufacturer has different in algorithmic detail optimization, each has its own characteristics. For example, H series is adopted for all series of million HD network cameras. 264 compression algorithm, can support 720P/1080p full frame rate high definition video output. While ensuring high-quality image quality, the H. The 264 technical standard has obviously improved and improved the compression coding efficiency. The compressed data volume is only 1/3 of MPEG4, which greatly reduces the transmission bandwidth and back-end storage pressure. And considering different network application environments, H. The 264 compression algorithm can use different transmission rates according to different network environments, and can adjust the occupancy rate of bandwidth by itself. While optimizing network bandwidth resources, it also provides users with more practical application experience. In addition, HD network cameras support H. 264/MJPEG/JPEG video encoding supports dual stream. Users can perform multiplexing of video data according to actual application requirements, thus achieving multiple choices of application operations.

Compared with SDI cameras, both SDI cameras and HD network cameras can achieve 720P/1080P video quality. Even SDI cameras can provide higher transmission frame rates. Generally, they can achieve real-time image acquisition of 30-60 frames per second. . Moreover, SDI cameras have higher definition images, higher resolution, and richer colors, which can meet the high-definition monitoring needs of users. At present, as a surveillance equipment specifically for the high-end security market, SDI cameras have a relatively good application prospect in some special industries because of their superior quality in image quality and relatively simple system architecture.

However, SDI cameras and HD network cameras have room for improvement in terms of basic functions such as low illumination and wide dynamics. The SDI camera now uses more CCD imaging chips. The CCD sensor has better sensation than CMOS. Relatively speaking, the monitoring effect at night is better. The photosensitivity of the existing high-definition image sensor is not ideal, and the image captured in the dark scene at night is not good. Therefore, it is very important to effectively improve the performance of the image sensor. In addition to improving the low-light performance of the light-sensitive chips, infrared technology, thermal imaging and other technologies are now used to obtain the desired image effects, so that the night-time monitoring effect is improved, making the 24-hour monitoring more perfect and more efficient.

In addition, the most important thing is that the transmission distance of SDI cameras is very limited, and it can only transmit about 100 meters. The network camera can transmit only by relying on the network, which is the biggest difference between the two. Not only that, because of the large amount of data collected by the front-end SDI camera, its transmission process is not compressed by the codec, posing a great challenge for data storage. However, HD network cameras use H. 264 compression algorithm can save transmission bandwidth and storage space.

In summary, the author believes that SDI cameras and network HD cameras have their own application areas and focuses. They are complementary rather than replacing relationships. If the SDI camera wants to obtain a large application space, the transmission distance may be a bottleneck. Therefore, the transmission distance of the SDI camera needs to be broken. Only by continuously satisfying the needs of users, can the SDI camera have a broader development prospect. For the re-optimization of the network camera video compression algorithm, both high picture quality and low bit rate are still the focus of its performance improvement. Then, whether it is an SDI camera or a network camera (SD or HD), the basic high performance of low-lightness, wide-dynamic cameras, etc., needs to be continuously improved.

Intelligent back-end smarter network cameras currently include two solutions. One is to perform intelligent analysis at the front-end camera, and the other is to perform intelligent analysis on the back-end server. The advantage of front-end analysis is that it does not require server support and can be applied independently. The disadvantage is that the front-end chip processing capability is limited, and it is not suitable for complex, large-volume analysis. Especially in some special applications or single-function applications, such as traffic, flow statistics, etc., often only need to analyze the license plate information and count, etc. At this time, the application of intelligent analysis in the front-end camera, it can meet the user's monitoring demand.

Taking into account various factors, some manufacturers gradually introduced a unique solution that is to do analysis in the back-end, the advantage of the back-end to do analysis is strong processing capacity, and another advantage is more flexible, you can choose at any time Front-end analysis or no analysis. For video surveillance cameras used in large-scale shopping malls, real-time monitoring of shopping areas, elevators, passageways, entrances, exits, parking lots, and warehouses is needed. Because there are many monitoring points, image information redundancy tends to occur. One problem, some manufacturers have chosen to do intelligent analysis on the back-end servers. Users can select the key areas that the front-end needs to monitor for intelligent analysis at any time according to their own needs, so that the application has great flexibility, and the server has powerful processing capabilities. Can also deal with the analysis of large amounts of data, especially for users with front-end analog cameras, it is more important. The original analog camera can be preserved, and only the intelligent analysis module is embedded on the back-end server, thereby saving the cost of replacing the front-end to the network camera. It can be seen that the back-end intelligent solutions are widely supported by users in the industry because of their professional safety and reliability.

At present, from the point of view of China's entire security industry, network cameras have not been able to gain widespread popularity. The author believes there are three main reasons.

First of all, compared with the traditional analog monitoring system, which is very mature in technology, its networking and debugging are relatively unfamiliar to the engineering personnel who are familiar with the simulation system. As we all know, network cameras rely on the existence of the network, and the condition of the network itself is closely related to the application effect of network cameras. Looking at the impact of the basic network environment on the application, if the network monitoring system is based on an established network, whether the original network infrastructure can meet the monitoring requirements is a consideration; however, for a network monitoring system based on a new network, There will be relatively large investments in building a new network environment, and large investments may make it difficult for users to afford. In addition, the network camera's visit will be plagued by domain names, IP addresses and other issues, and network security issues are also the focus of the user's attention, so the combined effect of many issues have restricted the wide application of network cameras. However, in terms of network security, today's millions of high-definition network cameras have built-in internationally-standardized VPN protocols that can seamlessly connect with network security devices and can be securely connected to internal LANs through the Internet, thereby greatly improving audio and video. Security can prevent intrusion into the internal network through the network camera's network cable. In addition, generally, high-definition network cameras also have built-in UPNP protocol, plug and play, no need to change the settings of network security devices such as firewalls, also reduces the difficulty of debugging.

Secondly, some SD network cameras do not have the same image effects as analog cameras, and the network cameras also have troublesome non-real-time features, all of which have also boosted the analog cameras still being used on a large scale, and have firmly occupied the mainstream of the surveillance market. However, with the improvement of technology and the improvement of user needs, high definition has become a constant pursuit of security companies, and 720P/1080P high definition cameras are becoming more and more popular with users. Taking some mainstream manufacturers as an example, the entire series of millions of high-definition cameras launched last year were all H-compliant. 264 algorithm standard, can achieve 1920×1080, 1600×1200, 1280×720 full real-time, million HD, supports 1600×1200 video multi-stream video output, can easily achieve high-definition monitoring, can capture image details in real and clear, picture quality Compared with the traditional D1, the application area of ​​high-definition cameras is increasingly widened. It can be optimistically forecasted. Under the trend of high-definition and networked video surveillance, HD network cameras will show a rapid development trend.

Finally, the cost of webcams is still high. Compared with analog cameras, because the network camera market is not yet mature, its high prices make it prohibitive for many industry users. With the gradual mastery of the camera core technology in China's domestic large security companies, the domestic application of compression chips has been widely used, and many domestic security companies have launched a wide variety of network cameras. Some manufacturers also have a full range of high-definition network cameras that can provide users with a richer choice. This means that prices will inevitably have a significant downward adjustment. The current price of high-definition network cameras has increasingly approached traditional analog cameras, enabling more users to use network cameras.

Conclusion In summary, the understanding of network cameras is further deepened from compression algorithms, intelligent analysis and other technical aspects. In the age of major security, with China's safe cities, intelligent transportation and other projects in full swing, network cameras are gradually using their excellent performance in high-definition, transmission and other aspects, gradually formed the scale of application. We believe that under the continuous change of network camera technology, network cameras will break through their own technical constraints and the constraints of external conditions, showing a broad development prospects. It can be pleased to see that network cameras have become the mainstream direction in leading the development of video surveillance, and they will explode huge “energy” in more areas to improve their attitude and provide users with more professional and intimate reality. Application experience to achieve new changes in the human intelligent lifestyle.