Digital technology allows intelligent monitoring beyond analog monitoring system

Simulation system constitutes the current most video security systems, but these techniques based on traditional TV cable system has many drawbacks: construction and maintenance of the available connections broadband cable infrastructure costly; rely on human monitoring multiple camera to detect suspicious activity consumes a large number of people can monitor system to efficiently run for at least 5 kinds of work force, but also because of human error can cause false alarms and reduce security; most importantly, breaches of security and found this behavior of interval often too long and will not be able to prevent crime or even unable to catch criminals.

It is no exaggeration to say that analog video monitoring technology has been a blind alley.

A lot of people need to sit there and stare at a screen, it is possible to produce a lot of false alarms and security also is impossible. Video surveillance market is facing a subversive of the technology, it relies on the concept of embedding the network machine intelligence to improve performance and reduce costs. New schema changes are drifting, intelligent will go out of the control center, go to the security system, or even eventually went into the camera.

This blueprint technology is an integrated digital system, it uses a series of advanced software, including image processing, video compression and rule-based machine intelligence, etc.

All these programs are complex and require a large amount of computation algorithm, run these algorithms need affordable, powerful DSP.

Decentralized intelligent

Intelligent video surveillance system uses many of the traditional analogue system to the same components, such as cameras, router network, video Server and monitor.

However, impersonation, and intelligent video surveillance similarities between also. The following figure shows a typical smart video surveillance system.

Through the network of strategic elements added to the smart technology, and use this intelligent technology, according to the network transmission of data for decision-making can significantly improve the performance of the video surveillance system.

Following a few key performance requirements.

Video collection: camera embedded intelligence, you can identify the object class, and then move through the shakes and zoom operation (Pan and zoom) provides more detailed information;

Data processing: the trend of digital data is very clear.

Next important direction of development is to millions of bytes of the original video is only the data that contains important information;

Distribution: intelligent safety network bandwidth is relatively small, because the transfer of only a few thousand bytes of data, instead of millions of bytes in video.

This network also without manual intervention will be routed to the correct network nodes;

Decision: the smart system to security personnel through a simple point-and-click interface to create, based on the rules of arithmetic processing related data;

Action and execution: the Elimination of redundant and useless information, as far as possible reduce the flow of data, which enables security personnel to patrol monitoring system by mobile devices, thereby enabling them to be closer to the scene, quickly taking the most effective measures;

Archive: Smart Security can also be data archiving and indexing for later forensic use.

¡Ø typical intelligent video surveillance system

Digital IP camera

IP camera compared with traditional analog camera has two advantages, they have the digital nature of the algorithm is simple and straightforward, followed by the IP camera also can make use of existing IP network, which eliminates the need for security technology private foundation Central Office.

However, this is just the beginning.

When the camera frame internal smart system integrates smart security software and high-performance DSP, these two magical effect also has another advantage.

Application software (such as the Object of the Onboard Video company) with edge detection and other imaging algorithms to create the monitoring area of an object's unique statistics file (statistical profile).

These algorithms can generate is called meta data information. Inference engine can then be used in artificial intelligence algorithm analysis related metadata to identify in the monitoring area of the object (or static objects).

In General, the target object category including person, non-human beings (animals) and vehicles.

You can also walk, run, identification of rambling and climbing, and other actions. Through the use of other algorithms can be defined to describe the object type and the actions of the other categories.

Additional steps are required, use rules-based decision making instruction camera shake, zoom and focus, and so on, or, if you move in a particular direction of someone who passed the surveillance zone virtual "trip network", the camera can even alert security personnel.

IP camera can use the above data, without the need for manual intervention.

It can be "smart" and select the most appropriate related conditions, for video acquisition and transmission of compression technology. For example, if an object is identified, the camera can close the trespasser's face, take a high-definition photos, the photos are compressed into JPEG files, and then sent to the appropriate storage area, or the breach of the rules require immediate response to send the file to the security personnel of the PDA.

IP camera will continue to use MPEG4 low-resolution video format for continuous record of the entire site, to store and subsequently used for forensic analysis.

The system saves the MPEG data for later use, this is because smart monitoring technology also has the important advantage of criminal investigation, it can save through analysis of data to search for specific events or time stamp.

Security analysts are usually based on new information check saved videos, such as in hundreds of kilometres away from non-related site collection to the unique identification of a suspect.

Use intelligent monitoring software and is used to perform the above algorithm for high performance, programmable DSP,The user can create one or more of the new rules, such as identifying a back with a special flag backpack suspects. Use a saved video generated metadata can receive new rules check so that the contrast is the suspect in the video and other unrelated field of suspect similar?? The entire process without requiring users to see and watch the video file.

Embedded control system for smart technology can also be other ways to optimize security.

Traditional video monitoring system of false alarm rate as high as 40 per cent security personnel will be wasted countless hours of time. Intelligent monitoring system of false alarm rate of only 2%, because it can be set very granular alert trigger rules. For example, in the parking lot surveillance process, you can stroll to the cause of the alarm, and passing a person does not trigger the alarm.

Network video server and router

In the IP camera smart technology embedded in and through the safety net send compressed data or metadata can achieve efficient system.

This means in particular can achieve low-cost new installation. However, to add an existing system that features high cost, especially if you need to connect the analog camera network.

By implementing the so-called "back end" cameras, network video server can resolve the problem.

In order to give traditional analog video surveillance systems integration mentioned above, can intelligent network video server installation system running DSP and software suite.

Network video server must have sufficient speed and efficiency, in order to handle multiple camera input signals, at the same time through the network for digital media content digitized, compressed, and network transmission, etc.

Remote user configuration and control is an important feature. Without a DVR, the server will also be able to save the video.

The primary task of the network server is more than one analog video stream into a digital stream.

After completing the conversion, the system that can extract metadata, the implementation of the relevant rules and perform the actions described above.

Network router implementation used to supplement the video capabilities of the server.

They can use metadata and a series of algorithms determine the system of the next operation. For example, you could describe as IP camera related directly to the security personnel. However, this is not easy, because the cell phone, PDA, laptop, and workstation in bandwidth capacity and acceptable format.

In router decide to send a JPEG file, generally you must receive equipment adjustment related data rate, and perform related information code conversion.

To do this, it needs to access the query table to determine whether the device supports H.26 received? or MPEG4 (or also supports both). As with other features, the router also has to maximize bandwidth utilization of other advantages.

Digital camera

Digital camera (DVR) is an intelligent monitoring system, an important part because it is able to provide efficiency and access performance is much higher than that of video storage media.

Compression technology allows for smaller spaces to store more data, and its advantages are not limited to this.

In addition, DVR also has perfect "suspended" in the still image function, this function is commonly used to view the video must be constructed in security applications.

DVR can either be used as a stand-alone device can also be used as a PC card.

DVR saved Digital Imaging does over time reduce quality and allow random access; most importantly, you can access the machine intelligence and conversion-related information.

DSP can be used in accordance with a variety of standard and non-standard video formats to compress images. As a global leader in DSP design and vendor, Texas instruments (TI) and its partners are already perfect intelligent image analysis functions, as well as a variety of network protocols support has made great efforts.

Signal processing

Intelligent monitoring system to efficiently run for at least 5 algorithm: image processing (edge detection and other object recognition technology), compression and decompression, for a variety of digital codec ADCs and code conversion, video data to private metadata conversion and is used to transfer data in the network of agreements.

Development of low-cost system is the best way in programmable DSP chip implementation of the required processing power and flexibility.

Another prerequisite is to provide suitable hardware platform of signal processing algorithms.

Then you need to consider the system of control.

In the schema, the DSP is the implementation of signal processing algorithms. Need to be controlled on the network, but this usually requires MCU. Table 1 gives the DSP and the MCU in advanced intelligent monitoring system requires collaborative implementation of key features.

To the total cost of the actual control system, the processor can be equipped with a DSP and MCU core, however, this also depends on the relevant application.

Although single-processor chips have been sufficient to meet present many applications, however, dual-core processors can significantly cut costs and improve performance.

Digital technology beyond artificial monitoring

Over the past 10 years of various dinosaurs??

Hammer has enabled Government and enterprise security as a top priority. With video monitoring range expanding, engineers and managers have started on the current simulation implementation concerns. In addition, for performance and budgetary reasons, use of cable networks, through attended transfer of command and control center is also a video to attract more and more critical.

Leverage software intelligence on the data from the video gets parsed in monitoring system at all levels have the advantage.

In addition, we make the monitor peripherals with the highest possible degree of intelligence (such as increase the camera's intelligent degree) will help to reduce the processing load, improving information quality and accelerate the speed to take remedial measures.

Additionally, with DVR storage data and video so that we can in the event of a security problem for weeks, even months after a subsequent forensic analysis.

Because intelligent video surveillance application FengRich and varied and has a very obvious price/performance advantages, so that it is a kind of subversive techniques not extreme.