Since the development of IP intelligent video surveillance, it has brought tremendous changes to the video surveillance industry, injecting tremendous vitality into the video surveillance industry. Since the construction of video surveillance is no longer limited to one floor and one campus, video surveillance has broken through the restrictions of geographical, space, and scale. It can be said that where there is a network now, there is video surveillance. The deployment scale of video surveillance has also expanded from dozens of roads to hundreds, thousands, and tens of thousands of roads. The deployment model has also grown from small to large.

The rapid expansion of the construction scale has also brought a series of problems. How to transfer massive data, how to display, how to store, how to manage, how to extract? So intelligent monitoring came into being. For some people, intelligent monitoring is intelligent analysis. It seems to me that this is a misunderstanding. Not to mention that intelligent analysis is still in the initial stage of technological development, and its accuracy and practicality are criticized. Even if one day develops mature enough, intelligent analysis can only solve the extraction of huge amounts of data, and other problems cannot be improved. Therefore, I think that any single monitoring platform function can only make minor improvements to the problems brought about by massive monitoring data, and often solving a problem brings a new problem. Only by using systematized smart technologies can we better improve and resolve these conflicts. Next, we will discuss the problems brought about by the massive video data and the solutions that the H3C IMOS intelligent monitoring platform uses with system-based smart technologies.

Question one: How to effectively monitor the current video surveillance, at times hundreds of thousands of roads, clients and monitors up to several dozen, the monitoring staff to traverse all the images become impossible tasks. Especially in the era of high-definition, with high-definition monitors, high cost scary; with the PC client, PC performance is a bottleneck, a client can see a few ways 720P, 1080P images? Three or four way, the client's multi-screen basically became a display.

The H3C IMOS intelligent monitoring platform solves this series of problems through systematic intelligent video technology. The entire intelligent video technology system consists of several key technologies and functions: virtual resources, group planning, intelligent analysis, linkage planning, and video intelligent processing. Through this series of technologies, effective information is highlighted, monitoring efficiency is improved, and loss of decoding resources is increased.

Virtual resources, through permission control, resource allocation and a series of functions, group a large number of camera image resources and allow monitoring personnel to do targeted supervision. For a monitoring person, thousands of cameras are reduced to dozens of hundreds of monitoring objects. The monitoring of dozens of images of hundreds of roads is still very high for monitoring individual personnel and maintaining effective monitoring. Then through the group plan, pair the camera image with the monitor, monitor client pane, and set the handover plan. Batch cameras can perform batch polling at the corresponding decoder, greatly improving the efficiency of surveillance video surveillance.

Through these two key functions, the monitoring personnel's workload is basically reduced to a minimum. However, the problem still exists. Because no one can maintain a high degree of concentration for a long time, there will always be omissions by the monitoring staff. Therefore, the two technical functions of analyzing and linking up plans can be used to find out what is missing, and to ensure that key images are not missed.

Through the foregoing introduction, we can see that the efficiency and effectiveness of monitoring personnel are related to the number of ways to monitor at the same time. Within a reasonable range, the more channels that are monitored at the same time, the higher the efficiency. Due to the high consumption of high-definition images on PCs, a client can solve a very small number of HD images. In fact, in multi-split screen mode, the playback effect of any image is far from the HD effect. IMOS intelligent monitoring platform, through video intelligent processing technology, automatically calls low-stream low-resolution SD images from the monitoring HD front-end in multi-split screens, and automatically switches to high-rate HD images in single-channel decoding. This greatly reduces decoding loss and improves decoding efficiency.

Question 2: How to ensure the integrity and security of video data The greatest value of video surveillance lies in the timely detection and afterwards forensics. The key to obtaining evidence afterwards is how the video is stored completely. IP Intelligent Video Surveillance has a powerful video feature that looks like a giant. The problem is that IP surveillance video, from the front-end coding to the final storage is completed, through the network transmission, storage management server, storage devices, hard disk write multiple links, any one link will become the "giant" weakness.

Therefore, a mature IP intelligent monitoring platform requires perfection in all aspects of video data transmission and management. No one can guarantee that the hardware never fails. This requires the IP intelligent monitoring platform to have perfect video data processing technology. To solve these problems, the H3C IMOS intelligent monitoring platform introduced a multi-dimensional intelligent storage system. The multi-dimensional intelligent storage system greatly enhances the robustness of all aspects of video storage through technologies such as buffering, direct storage of iSCSI blocks, N+M backup, and MAS. With the use of encoding device front-end buffering and automatic recording technology, once the network is restored, the video data in the cache will be automatically recorded to the central storage, which solves the problem of video data loss when the network fails.

Adoption of iSCSI Block Direct Memory Technology in Video Surveillance System This technology is the core technology of multidimensional intelligent storage systems. The mature application of the iSCSI block direct storage technology solves a series of problems encountered by ordinary IP video surveillance massive video storage. First of all, compared to ordinary file storage, the efficiency is greatly improved. All video data is stored in blocks and read in blocks, which reduces the performance loss caused by file packaging and file unpacking. Second, since the file storage itself is a video The video data is packaged as a complete video file that can be directly downloaded, copied, and played. With iSCSI technology, all bare data blocks are stored, so they cannot be directly downloaded, copied, or played. This greatly improves the security of video data and prevents illegal transmission and illegal use of video data. The most important thing is that the iSCSI blocks are directly stored, and the video data is directly stored in the IP SAN from the front-end encoding device. Without any processing or forwarding by the server, the efficiency of the iSCSI is greatly improved, and video data resulting from the failure of the monitoring platform server is completely resolved. Lost problem.

Application of storage device N+M backup technology We know that the failure of the storage device is catastrophic for the IP monitoring platform, which will result in the loss of many surveillance front-end recordings. The conventional multiple storage technologies, although solving this problem, greatly increase the construction cost of the IP video surveillance platform due to the need for twice or more times the number of storage devices in the conventional situation. The N+M backup technology requires only a small amount of backup storage devices for the N primary storage devices. Once the primary storage device fails, the video is automatically stored to the backup device. Therefore, the N+M backup technology not only solves the problem well, but also greatly reduces the construction cost.

MAS storage optimization MAS storage optimization consists of a series of optimization algorithms, including: Cache optimization, RAID reconstruction algorithm optimization, RAID anti-disk error kick optimization algorithm, RAID5 optimization algorithm, etc., greatly improving the security of monitoring hard disks, array groups, and effectiveness.

Question 3: How to make the video data information We know that video data is massive data, but in the massive data, really useful video is very few. It can be said that a monitoring platform, only no more than 1% of the video data is valid information. Once it is necessary to collect evidence afterwards, finding effective information requires a lot of manpower to take a needle in a haystack. This requires the intelligent monitoring platform to have the ability to quickly locate, effectively archive, and facilitate viewing. The H3C IMOS intelligent monitoring platform provides a series of video intelligent extraction technologies and functions through a video intelligent forensic system, which greatly improves the efficiency of the effective use of video information. In fact, the core technology of the intelligent forensic system is the direct storage of iSCSI blocks.

The application of video equalization and positioning technology automatically divides a certain segment of video into segments and plays them synchronously, which greatly improves the efficiency of the positioning of effective video information. Imagine a 24-hour video that can be equally divided into 150-second videos by doing two 24-hour positioning. This will allow our surveillance staff to find effective video information for the day in only two or three minutes. It is because of the iSCSI block direct storage technology that can guarantee accurate positioning of video data for any length of time. The error rate is much lower than the normal file storage.

Application of video archiving technology Video archiving is divided into two parts: archiving of video data and archiving of video and intelligence information. The recording of video data is completed by video backup. There are not only manual and scheduled backup functions, but also alarm linkage backup. The application of alarm linkage backup realizes the intelligentization of effective video data archiving through intelligent analysis or conventional security alarm linkage video backup, without manual intervention. The video and intelligence information archiving, through the use of pre-recorded video effective data to do bookmarking, case information description, making important video can be repeated quickly query, reuse.

Application of video integrated viewing technology This technology includes several functions such as second-level playback of video data, multiple-channel simultaneous playback, instant playback, and full-frame playback at multiple speeds. Multi-channel synchronous playback allows the monitoring personnel to accurately play the video of the multi-channel associated monitoring point. By surrounding information, it can better find effective video information. In real-time monitoring, in real-time monitoring, once the sudden situation occurs, it can directly Return live images and switch to video playback automatically. Playback at full frame speed ensures that all video information can be viewed when viewing multiple times. Usually when doing multiple multiple playback, it is achieved by skipping frames, which will result in the loss of a large amount of video information.

Question 4: How to control the monitoring of network traffic Discuss how to effectively use IP intelligent monitoring, and then focus on how to control the monitoring of network traffic. The network of IP intelligent monitoring is a big issue. The distribution of traffic in the monitoring network is very unpredictable and sudden. The IP intelligent monitoring platform usually adopts unicast networking and streaming media networking solutions. There are two fate problems in this solution: the single-point bottleneck of the streaming media server and the high concentration of traffic on the network node where the streaming media server is located. In fact, these two problems can be solved with the same technology—multicast technology. We know that the most important role of streaming media servers is to distribute and distribute media streams. Now, multicast replication is implemented by most of the switches themselves. In the case of multicast replication, traffic is distributed at the edge of the decoding device. Switch ports do not have the problem of highly concentrated traffic on a network node. Therefore, the conventional streaming media technology not only wastes resources but also brings significant problems to the networking.

Of course, streaming media servers are not useless, depending on how they are used. In a monitoring network, there may be some old and low-end network devices that do not support multicast and cannot implement network-wide multicast. At this time, multicast media is converted to unicast through a multicast conversion technology on the streaming media server. This provides a live monitoring and streaming service for clients and decoding devices under these network devices. This problem can be solved. However, as long as the streaming media server is used, the traffic of the peripheral network of the streaming media server may increase dramatically. Often, the front-end encoding device and the decoding end are in the same network, but it needs video streaming to move from the edge of the network to the central streaming media server to the edge of the network. This greatly increases the traffic pressure of the mainline network. If the streaming media server is deployed near the network at each decoding end, the construction cost will be greatly increased.

The H3C IMOS monitoring platform solves this problem through intelligent routing technology. Through intelligent routing technology, the decoder first obtains the video stream directly from the near-end encoding device, then selects the streaming media server that is closest to the network path to obtain the video stream, and finally chooses to obtain the video stream from the streaming media server that is far away from the network path. This ensures optimal distribution of network traffic.

The summary lists so many issues and introduces so many solutions. I believe that this industry will encounter more problems and users will also put forward higher requirements for this industry. Every technology, every product, and every industry is growing and growing in such a process. The key is to solve the problem of how to improve the product. For high-speed growth and long-term IP intelligent monitoring products, simply whether or not a certain function is available to promote the growth of the platform can only be a supplement to a platform, and cannot really bring usability. Only systematically combining the business needs of users and adopting systematic and intelligent technologies can truly provide users with a powerful and powerful IP intelligent monitoring system.