Application of intelligent video analysis technology in security field

Tag: Video Surveillance Pattern Recognition

This article refers to the address: http://

The development trend of modern network video surveillance systems is large-scale networking, distributed deployment, and intelligent monitoring. Intelligent video analysis based on computer vision combines technologies in multiple subject areas such as image processing, pattern recognition, artificial intelligence, automatic control and computer science. Compared with the traditional video surveillance system, the intelligent video surveillance system can analyze and mine valuable information from the original video, change the manual servo to active identification, analyze the event and analyze the alarm and alarm.

Intelligent Video Surveillance (IVS: Intelligent Video Surveillance) analyzes the video image content of the surveillance scene based on computer vision technology, extracts key information from the scene, generates high-level semantic understanding, and forms a monitoring method for corresponding events and alarms. If the camera is regarded as the human eye, the intelligent video surveillance system can be understood as the human brain. Intelligent video surveillance technology often uses the powerful computing function of the processor chip to perform high-speed analysis of massive data in the video picture, filtering out information that the user does not care about, and only providing useful key information for the monitor.

The emergence of video intelligence analysis and its characteristics

There are two basic problems to be solved in the intelligent video system: one is to free the security operators from the complicated and boring "screen-to-screen" task, and the machine to complete this part of the work; the other is to quickly in the massive video data. Search for the image you are looking for. According to statistical analysis, the security operator will miss 90% of the video information after staring at the screen TV wall for more than 10 minutes, making the work meaningless. In the London Underground bombing, security personnel spent 70 hours working to find the information they needed on a large number of tapes. Therefore, based on the above two points, the choice of video analysis system will be able to free people from heavy labor, thereby improving efficiency, intelligent video analysis system will become a core component of video surveillance system in the future.

The background generated by intelligent video analysis comes from the most basic needs. For example, when security guards face hundreds of cameras, they cannot really prevent or interfere when risks occur. Most of them rely on playback-related videos for post-processing; Some non-security applications, such as business flow statistics, target recognition (license plates, faces), etc., also require automatic intelligent statistical recognition. Intelligent video analysis transfers the analysis and identification of events to a computer or chip, freeing the duty personnel from the work of the "dead stare" monitor. When the computer finds a problem, an alarm is generated and the attendant performs the corresponding processing.

The main advantages of intelligent video surveillance:

• Fast response time: millisecond alarms trigger reaction time;

• More effective surveillance: Security personnel only need to pay attention to relevant information;

• Powerful data retrieval and analysis capabilities: Provide fast response times and survey times.

Motion detection is the foundation

Most intelligent video analysis is based on moving target detection technology. Firstly, the intelligent analysis system can accurately detect the moving target, effectively separate the moving object from the image background, and extract the moving target information.

From the practical application of computer vision, the main challenges and problems to be solved in moving target detection and recognition and analysis can be summarized into three aspects, namely, the robustness, accuracy and real-time of the algorithm.

Robustness

Robustness is the robustness of the system to characterize the insensitivity of the control system to characteristics or parameter perturbations. The robustness of the moving target detection algorithm is to enable continuous, stable detection, analysis and identification of moving targets under various environmental conditions.

The main reasons that affect the robustness of the algorithm are as follows: changes in target state, changes in ambient illumination, irregular deformation of the target caused by partial occlusion, and temporary disappearance of moving targets caused by total occlusion.

accuracy

Moving target detection and recognition for different application situations, the detection and recognition rate is different, almost impossible to achieve 100% detection success, that is, there are false detection and missed detection. Because the actual monitoring scene environment is complex and ever-changing, there are a lot of noise and interference situations. The optimization of the algorithm can improve the accuracy of certain detection. At the same time, it can only be based on the actual demand, the false detection rate (false alarm rate) and missed detection. A balance between the rate (missing rate) is sought.

real-time

A practical intelligent video surveillance system must have the ability to process video image sequences in real time. Since the processing method of the video moving image is based on the processing of the two-dimensional digital signal, the processed object contains a huge amount of data and information, and the algorithm cannot be calculated too complicated, and must be fast and real-time. For real-time analysis of early warning tasks, computational complexity is critical in order to allocate more resources to more advanced tasks. Among them, real-time and robustness are often contradictory. How to seek balanced development is the key to technology.

In particular, it has been developed by the company's self-developed video motion detection algorithm, which is less affected by light and lens jitter than the classic motion detection algorithm, and has less computation time, which is more suitable for real-time product development. At the same time, based on the autonomous algorithm, the package development kit (SDK) is integrated in the company's digital security system software platform and network camera series products, achieving reliable application at the system level and product level. At the same time, the project team is extensively building video libraries for a variety of scenarios, using video algorithms for testing multiple scenarios to better improve the algorithm and reduce the number of parameters that need to be adjusted to better meet actual application needs. .

Application of intelligent video analysis system

Companies engaged in security and video surveillance system integration and product development for many years have introduced intelligent analysis into video surveillance while further maturing traditional surveillance technologies. At present, the intelligent analysis of video has been applied to monitoring systems in the power industry and safe city.

Substation video surveillance system for the power industry

At present, video surveillance systems in the power industry are generally subject to post-event processing, which is often too late. Therefore, in order to prevent it from happening, the perimeter can be prevented within a certain range in the substation. When a suspicious person is found to invade or enter the warning line, the monitoring system is required to automatically detect the intrusion target, identify the intrusion trajectory, and issue an alarm notification management. The staff went to deal with it.

Set all prohibited intrusion areas within the perimeter monitoring range as zones. In this way, when an intruder invades the defense area, the intelligent monitoring system automatically locks and identifies the action track, and issues an alarm. The alarm here is divided into front end and back end. The front end alarm can be realized by the sound and light alarm. When the intruder triggers the alarm, the monitoring system will sound an alarm to warn the intruder and open the strong light to prevent the intruder from hiding. The back-end alarm is implemented in the system software platform application, and the management personnel handle the intrusion behavior, so that “pre-prevention” can be achieved to avoid losses.

Safe City Monitoring System

For the safe city monitoring system, one aspect is mainly reflected in some important road sections, communities, public places, etc., to monitor and alarm the suspicious targets appearing through video surveillance. On the other hand, it focuses on the post-operation management process of the monitoring system to detect the common faults of the front-end camera and video image quality through video analysis technology, and realize the effective maintenance of the monitoring system.

The demand for video surveillance in Ping'an City is complex and the system capacity is huge. It involves not only traffic vehicles, personnel gathering monitoring and violation alarms, monitoring of illegal traffic such as illegal parking, but also the appearance of suspicious people in the community and the alarms of staying, etc.; To combine the Internet of Things and cloud computing technology, build a massive video storage and content analysis and retrieval system.

Intelligent analysis results planning

At present, security and video surveillance system integration and product development companies have successfully implemented some of the video intelligence analysis results: intrusion detection, over-the-line detection, remnant detection, robust testing in some typical complex scenarios; planning implementation: personnel aggregation And personnel statistics, fire and smoke detection and alarms, etc.

• Intrusion Detection: Given a virtual forbidden zone, if a moving object invades into the restricted zone, an alarm is generated;

• Over-line detection: Given a virtual line, if a moving object crosses this virtual line, it will alarm;

• Remnant detection: An alarm occurs when an object (such as a box, parcel, vehicle, person, etc.) stays in a sensitive area for an extended period of time or exceeds a predefined length of time. Typical application scenarios include airports, train stations, subway stations, etc.

• Number of people: Count the number of people or objects that cross the entrance or designated area. For example, for the owner to calculate the number of customers who visit their store one day;

• Smoke detection: On the basis of realizing the detection of moving targets, it can be judged that the moving target is smoke, so as to realize the alarm; the smoke is marked with a rectangular frame to realize the tracking of the smoke;

• Flame detection: On the basis of realizing the detection of moving targets, it can be judged that the moving target is a flame, thereby realizing an alarm; the flame is identified by a rectangular frame to achieve tracking of the flame.

Conclusion

With the continuous advancement of major projects such as “National Emergency Response System”, “Safe City”, “Safe Construction” and “Science and Technology Strengthening Police”, the domestic video surveillance market has good prospects. The application of intelligent video-related products will continue to expand from a relatively concentrated area to various industries as the size of the video surveillance market continues to grow. As far as market demand is concerned, the current market awareness of intelligent video analytics is constantly improving. Security and video surveillance system integration and product development company will rely on years of system integration and product development experience to fully understand the needs of different industry segments, and combine intelligent video analysis and detection technology with security monitoring industry to build reliable and practical intelligence. Analyze applications and contribute to the development of intelligent applications for security monitoring.

Control way-APP

The light color, brightness, cycle mode and timing can be controlled by App.The App are developed on the most advanced Bluetooth Mesh technology. [LinkupHome" App can be downloaded in the App store or Google Player, then you can control our product without any complicated steps. The App is stable, easy control and multifunctional.

Introduction

Smart down light with both warm light and RGB colorful light, it can meet the daily lighting and ambient lighting, high efficiency and energy saving, light source is stable and without strobe. Three color temperatures can be choosed for white lighting, 3000K, 4000K and 65000K. The color, brightness and RGB lighting auto-cycle mode can be controlled on the App. Three sizes of down lights for customer`s choice, small , middle and big sizes can meet different requirements, and the installation is very easy.

Product Parameters

Lamp Power: 9W
Working Voltage: AC110-240V
Frequency: 50-60HZ
Lumen: 805LM
Color: RGB+CCT
Protocol: Bluetooth Mesh
Wireless Transmission: 20m

Protocol

The advantages of Bluetooth Mesh are fast connect, low power consumption, no password required and Ad-Hoc Network. No need WIFI and hubs, as long as you have a Bluetooth-enabled smart phone, you can experience the smart light. When you install several smart down lights, the self-organizing network function can make the signals free connect, break the limitation of distance.

Big Size Downlight

Big Size Downlight,Large LED Downlights,Big Size Fire Rated Downlight,Big Size LED Downlight

Ningbo Homey Photoelectric Technology. Co., Ltd , https://www.linkuphome.com