Abstract: Intelligent video surveillance technology is a new type of video surveillance technology based on image processing and pattern recognition. In short, it is to find the moving objects in the image, track and analyze them, find "abnormal" behavior in time, trigger the alarm and take other measures to intervene.
Comparison of intelligent video analysis technology and existing monitoring technology
The video monitoring system of existing domestic ports is mainly composed of cameras, optical cables, matrix, hard disk recorders and TV walls. Due to the large number of video surveillance images and boring content, even if the existing system is equipped with staff on duty, in most cases it is still in a state where no one is watching. When a crime occurs, it is one of the main values ​​of the system to retrieve video playback and forensics from the hard disk recorder. Even if the duty officer is on duty, due to the physiological characteristics of the person, it is impossible to observe multiple images effectively for a long time, which may cause the omission of suspicious events and produce a wrong judgment on the safety form.
Intelligent video surveillance technology can be understood as using a computer to help the watchman "watch" the surveillance video. The high reliability of modern computers can provide 24 hours of uninterrupted protection. The problem of omissions caused by personnel fatigue is fundamentally eliminated. At the same time, it can prevent the possibility of collusion between the inside and outside of the monitoring personnel.
The existing intelligent video analysis technology can not replace people, it still needs people to identify the nature of the alarm, and then decide the disposal method. Its value lies in greatly reducing the work intensity of the monitoring personnel, so that the quality of monitoring has been qualitatively improved. Compared with traditional video surveillance systems, intelligent video analysis can detect and even prevent the occurrence of criminal behavior in time. This has obvious implications for sensitive and critical infrastructure such as ports.
Functions of intelligent video analysis technology and its application scope in ports
The demand for video surveillance systems in domestic ports mainly stems from the need for anti-theft, production safety and anti-terrorism. The valuable materials in the port and the monitoring of key key departments are more typical application scenarios.
Since China has signed the Sonas Convention, all ports exported to the United States must meet certain security standards. At the same time, as an important link in the national economy, the anti-terrorism of ports has also been included in the agenda of major domestic ports.
The main function
The intelligent video analysis systems currently on the market usually have the following functions:
1. Image acquisition / interface. The vast majority of intelligent video analysis algorithms are based on uncompressed image formats, such as RGB or YUV, so the image signal is sent directly to the video analysis unit without compression after being collected. Almost all video analysis systems have their own image acquisition function, usually through BNC input analog image signal. Image signals in existing image monitoring systems usually exist in the form of compressed image streams, which can be decompressed and restored to the original image format before analysis.
2. Detection of moving objects. Simply put, motion detection is to find moving objects in an image. Moving objects can be simply defined as the changing parts of the image. Some elementary motion detection algorithms are based on these concepts. The false alarm rate of such methods is too high to be suitable for real-time alarm systems.
Not all changes in the image are moving objects that we are interested in. For example, the changes introduced by the camera itself include pixel noise, overall brightness changes caused by the camera's automatic iris control circuit, and high and low frequency periodic noise signals introduced in image transmission. , Sudden changes caused by infrared camera periodic calibration. The changes introduced by the external environment include rapid changes in ground light in cloudy weather, shadows of moving objects, waves or sparkling on the water surface, swings of branches on land, halo caused by car headlights at night, rain and snow weather and other phenomena.
In addition, the camera is easy to shake on windy days, especially on high light poles. The image changes caused by these phenomena should be filtered out. They can be solved by algorithms or other technical means.
From the perspective of the algorithm, it can be simply divided into two categories. One type is to build a background model and find moving objects by comparing with the background model. The other type is to find moving objects through the "optical flow" method by discovering the effects of moving objects on the optical flow field. The other is a method between the two or a combination of the two.
3. Multi-object tracking. The essence of tracking is to string together the same objects found on each frame in chronological order. This field itself is a relatively independent and active research field. The main research direction is to effectively track under complex environments, such as multiple moving objects, multiple cameras, and mutual blocking, disappearance, and reproduction of moving objects.
In actual monitoring applications, especially in some intrusion alarm application cases, the requirements for the tracking algorithm are relatively low. Existing systems are not ideal for tracking the "fusion" of moving objects and other complex application scenarios. However, referring to the past technological development speed, this aspect will soon be improved.
4. Analysis of behavior characteristics. Behavioral feature analysis is to look for events that meet predetermined behavioral features from the image. Typical applications currently on the market include:
(1) Classification. Determine whether the moving object is a person, a car, a ship, or an airplane. Stop or suddenly accelerate, for example, when the vehicle breaks down in a tunnel or highway, and escapes after robbing on the street.
(2) Wandering. Such as people observing outside of sensitive areas. No warning to pedestrians and vehicles passing normally.
(3) Remnants. For example, place explosives at airports, oil depots and then leave.
(4) Missing items. For example, when the valuable exhibits of the museum are found to disappear, the system will immediately call the police.
(5) Number statistics. For example, statistics on the number of people entering supermarkets and other places, and combined with sales data to draw a curve of average consumption in a day.
(6) Crowd density. For example, when there are too many people, the police will be called, or when the crowd suddenly disperses and an abnormal situation occurs.
(7) The personnel fell to the ground. For example, when a person suddenly changes from standing to lying down.
In general, intelligent video analysis can do many things and requires effective communication between video analysis developers and end users. Because intelligent video analysis is still a relatively new technology, the circle that understands this technology in China is currently only extended to the level of integrators, and many application scenarios suitable for video analysis technology have yet to be developed by the market. But one thing is clear. Enterprises must master core technology and have independent research and development capabilities. The intelligent video analysis market is composed of many small market segments, and new applications are constantly emerging.
5. Set the alarm conditions. The introduction of "intelligence" in video surveillance has greatly enriched surveillance content. The currently available alarm elements include area, time period, type of object, size, direction of movement, speed, behavior characteristics and many other contents.
6. Alarm linkage. After the intelligent video analysis system finds abnormal conditions, it is usually necessary to verify the authenticity of the alarm, such as a detailed investigation of the alarm event by another gimbal camera, or timely notification and remind the monitoring personnel. Commonly used real-time prompt methods include voice, pop-up images, sending text messages, screenshots and other means to prompt the monitoring personnel.
special function
Intelligent video analysis can also address the special needs of ports.
Especially in safety production. For example, the illegal parking problem on the trunk road of the container yard, the inaccurate parking position of the container truck under the bridge crane, the spontaneous combustion alarm of the coal yard, etc. Since the intelligent video analysis technology itself can be applied to many fields, the communication between the port and the video analysis company can bring many new problem solving ideas to the port.
Special problems encountered
Compared with other intelligent video analysis application scenarios (such as prisons, power plants, banks, key institutions, etc.), the port project has the following difficulties that require attention:
1. Large range
The perimeter of the port is usually more than a few kilometers. It is usually quite empty. The available installation points may not be able to meet the requirements of a typical intelligent video analysis system. This is mainly because the typical detection distance of existing products is usually in the range of tens of meters, and individual manufacturers can do farther. Technically, it is mainly a contradiction between resolution and processing power. If the resolution is too high, some low-end systems cannot process such large amounts of image data in real time due to the increase in data volume. For ports, analysis using D1 (704 times 576) images is often encountered.
2. Camera shake
Since the port is usually open, facing the sea or river, camera shake is definitely present. Even if the camera is fixed on a very thick light pole, it can not avoid shaking in windy weather. For fixed cameras that use the background model method, camera shake must be effectively compensated before they can be used, otherwise a large number of false alarms will be generated.
3. The waves on the water surface are reflective.
The above technical difficulties are solved in some existing commercial systems. Therefore, it is necessary to screen when choosing a system.
When it is required to monitor the waters of the wharf, such as the alarm of the intrusive small sampan, if the sparkling phenomenon caused by the water surface waves and the reflection of sunlight cannot be handled well, it will also cause a large number of false alarms.
Technology development trend
Whether it is technology or market maturity, intelligent video analysis is at an early stage. Some new products and new functions are constantly emerging. In general, the foreign market is more widely used than the domestic market; there are fewer domestic independent research and development manufacturers, and more manufacturers integrate foreign systems. In the long run, the development of video surveillance systems in the direction of intelligence is gradually recognized by government departments, integrators, and end users. How to choose the time to enter the market, to achieve a balance between using intelligent video analysis technology to improve port security level, thereby improving cargo owner satisfaction and looking for mature technology, is a problem that port security departments need to consider based on their own circumstances.
Desktop LED Grow Light
Ddesktop led grow light, Led grow light garden, Led grow light homedepot, Led grow light walmart
Shenzhen Ri Yue Guang Hua Technology Co., Ltd. , https://www.ledlightinside.com