China's security intelligentization has a bright future

At present, the most talked about topic in China's security industry is “how to achieve high-definition”. In fact, video surveillance from analog to digital, then to the network, high-definition, and smart, has been solving two problems: "real-time monitoring" and "record storage." Intelligent video analysis solves the manual monitoring by the machine, and the other is quickly searching for the desired image in the massive video data. Therefore, the next security explosion point will surely be an intelligent application. In the future, perhaps intelligence will be the "standard" security system.

First, intelligence, the next node of security technology rush

Currently, surveillance systems often use digital cameras to collect data and rarely perform image processing. The digital content is then converted to an analog signal, transmitted over the existing infrastructure, and reached the command center before being converted back to a digital format for analysis and eventual storage. Throughout the process, it is difficult for us to deal with information. Intelligent security systems are very different, they are mainly end-to-end digital systems, using IP networks, because the flexibility of IP transmission is much higher than the traditional point-to-point analog transmission. Essentially, each camera becomes a smart webcam that can analyze video data at any point in the communication path and act on it. The intelligent video analysis process mainly consists of foreground detection, target detection, target tracking, target classification, trajectory analysis, and event detection. Each part has professional applications. However, in order to improve the performance of the algorithm in some scenarios, some additional modules are added to the algorithm, such as anti-jitter, shadow suppression, and headlight suppression modules. Therefore, smart algorithms often have multiple implementations. This is why the nominal functions of various smart products may be the same, but the performance difference may be large. A set of intelligent algorithms needs to consider various complex situations that may occur in the design. The algorithm modules must have adaptability to complex scenes, and add appropriate additional modules to improve the processing effect of the algorithm on the scene. In general, the products developed by professional intelligent video analysis manufacturers have the following three functions in a broad sense:

(a) Content analysis. In all cases, the number of important safety information is certainly much lower than the amount of background information, such as static information such as buildings and highways, or useless dynamic information such as disturbances or water waves. The equipment can eliminate useless information and focus on any important activity objects such as vehicles;

(b) Object identification. In order to successfully separate important safety information, equipment must be able to classify objects. For example, the device should be able to identify people, animals, vehicles, or other objects. From a software perspective, object recognition is accomplished by the content analysis engine;

(c) analysis of activities. Devices can identify important activities and compare them with a set of rules developed by security personnel. We may wish to give a simple example to illustrate how to formulate rules on breaking into safe areas. In order to implement this rule, security personnel use electronic pens or a mouse to draw a virtual network on the background video graphics to set a safety margin. . If an object identified as a person crosses the "network," the alert mechanism is activated and the software alerts or suggests. However, from the perspective of industry development needs and technological frontiers, intelligent video analysis can also develop the following two functions:

Generate important data

Generating metadata is a complex task that completely breaks the traditional concept of video. Because the video is to be monitored by a computer rather than a human, graphic enhancement must be performed. Computer analysis needs to improve contrast, which is a good example of a problem. Since motion detection and pattern recognition algorithms depend on the edge of the recognition object, it is important to improve the contrast of the image. We can build an edge sharpening algorithm in the camera subsystem. Tracking moving objects in the camera's field of vision, and comparing the activity of the object with a series of rules through the activity reference engine, often leads to conclusions and sometimes calls for safety actions. We can issue commands to the camera's positioning system to perform pan, zoom, and focus operations on the object or to command the camera to change direction to continue tracking the object beyond the original field of view.

Post-mortem forensic analysis

In addition to the ability to generate data to provide users with statistical analysis of security and management, intelligent video analysis can also solve broader security issues than illegal intrusion. In the past few years, video surveillance technology has become the technology of choice for many government and corporate sectors. The trend of transition from analog to digital systems is a general trend. However, it is generally agreed that the transfer of compressed video to manually controlled command centers is a backward strategy. Using metadata such as content analysis engine and inference engine to analyze the metadata extracted from the video has advantages at all levels of the security system. In addition, having the highest possible level of intelligence for monitoring peripherals (such as increasing the intelligence of the camera) will help reduce processing load, improve information quality, and speed up remedial actions.

Obviously, the intelligent video analytics that assemble thousands of pets will become the protagonist of security.

Second, the status quo of China's security intelligence: ideal fullness, realistic bone

Although it is a "pampering" technology in the market, the application of intelligent video analysis technology in China is not easy. Some of the world's professional IVS research vendors, such as the United States, ObjectVideo, Vidient; Israel's NICE, Mate, IOImage; Australia's IOmniscient and so on. These manufacturers have entered the Chinese market one after another, causing foreign companies to monopolize the domestic market of intelligent video analysis. In the end, because of the difference between its technology and China's national conditions, it cannot really penetrate the Chinese market. However, most of the domestic smart video analytics manufacturers can identify the behavioral features that are limited to specific situations. For example, target tracking, cross-border, counting, target loss, object left, etc., are mainly used in relatively fixed application scenarios such as detention centers, prisons, museums, warehouses, plant sites, and subway stations, and in monitoring environments targeting the social level, due to monitoring. The complexity and variability of the scenarios have led to a significant reduction in the stability of the monitoring. The intelligent monitoring products that use the accuracy of the alarm events as the evaluation criteria have also highlighted their application limitations. Therefore, the intelligent video analysis technology is in an appalling position. In an embarrassing situation, what if we look into where the inflection point of intelligentized large-scale application is? It should be said that in addition to the congenital advantages of the previous analysis of its own technology, it is necessary to solve the following bottlenecks. Perhaps, the sandstorms and towers, and the explosion of security intelligence are just around the corner.