Intelligent analysis technology will solve the digital video surveillance technology barriers (1)

With video monitoring key technology continues to develop, through the ubiquitous surveillance cameras will help security officers handling, display, evaluation, and storing large amounts of video content.

But how will the operator from burdensome monitoring work, the current obstacles to solving video surveillance technology. The accompanying video intelligence technology to better address this challenge.

For safety monitoring system, the main task is to require a low-cost manner, in hours, days, weeks or even months of time video image transmission, storage and search.

This, along with the continuous development of key technologies, and security monitoring industry is experiencing significant technological upgrading.

The development of the main problems: millions of pixels of low-cost digital camera are emerging; low-cost high-definition (HD) monitors, with compression coprocessor for the launch of a new DSP; IP technology popularization and consolidated site network convergence; power-over-Ethernet (PoE) technology to promote cost reduction, and no longer an offering for each camera is equipped with power; and mass storage device costs decrease, etc.

In addition, built-in calculation features of next-generation smart camera is already listed, you can do the video analytics, and assume a certain processing tasks, and operations staff from burdensome monitoring work.

These important technology development will contribute to the security personnel handling, display, evaluation, storing large amounts of video content, while the video is omnipresent surveillance camera network through various types of collected.

First, the bandwidth, latency, and storage capacity is still the major technical challenges

Safety technician is very clearly recognized that monitoring video and "Hollywood blockbusters"-style video completely different, not in the Studio.

Security surveillance cameras typically work in light dimmer, contrast is low, and the camera location typically capture the target area are not very satisfactory.

Luckily, safety monitoring system's main goal is, an object or a specific area, and therefore do not usually need to broadcast levels of image quality.

For high-end applications, the system may need a higher resolution to capture and read some information, such as the adverse environment to capture and read moving vehicle license plate number.

North American television standard frame rate to 30 frames per second, a source image resolution is 720 X 480 pixels.

Safety monitoring system resolution and frame rate can be far lower than the standard, because it doesn't need to capture all the details. So you have access to storage and channel density, etc. Therefore, many analog security surveillance system run frame rate only 7.5 FPS, image resolution is 352 X 288 pixels for (CIF).

By lowering the frame rate or resolution, to a certain bandwidth even to support four camera images, not just one.

If at the same time reduce the frame rate and resolution, under normal circumstances, a camera image takes up bandwidth can support up to 16 camera images. Of course, there are also save bandwidth costs, that is, the image quality degradation. In the frame rate for FPS, screen 7.5 may not smooth, the image resolution is 352 X 288 pixels, the details may rather ambiguous, however the operator typically can still make out the scene. In the traditional system, usually by lowering the frame rate or a resolution or at the lower of the two parameters to achieve the aim of reducing costs.

The new camera technology improved compression methods, and more low-cost storage device.

As a result, your system can support a higher frame rate and resolution, usually up to 720 X 480-pixel resolution and 30 FPS frame rate. If necessary, the new system can also be programmed with a lower frame rate or a resolution to reduce costs.

Security monitoring system in another key requirement is that the low-latency, delay involved in two aspects, one is the video appears in the display the time required, the second is the camera panning and zooming a suspicious event.

Real time video surveillance can sometimes due to delayed longer and affected, causing the joystick control and camera movements are not synchronized. Analog TV system before launching delay always than IP video system. In order to reduce the digital video delay, it should resolve Ethernet variable delay and loss problems, but also solve the video encoding and decoding itself caused delay. Typically, ensure good Web-based Central Office and adopt the advanced codecs and video compression technology, you can resolve these issues.

Video storage is also a challenge, because most security monitoring system must support storage capabilities, in days, weeks, months or even several years after the video content can also be transferred out of the retrieval.

Systems should be highly efficient compression algorithm and low-cost mass storage device, in order to meet the required amount of data storage. But considering the amount of data is very large, the storage technology should also be able to tag and can be retrieved quickly jump to a specific part, such as jump directly to someone, or a car leaving the car park and a frame.

This requires that the birth of the video analysis, computer analysis operations can be conducted to identify the target event and mark it down for artificial explanation or further action.

Second, depending on the application of different compression methods

As mentioned before, the compression scheme to achieve powerful video security monitoring system is essential.

Over the years, different compression schemes abound, but safety monitoring system has been mainly uses the MPEG and H.264 standards. These algorithms will frequently uses the reference image, subsequent images only stored relative to the reference image change of information, which helps to minimize the image dimensions.

Frame prediction is not always the best solution.

In the frame rate is low, the change between the different images may be larger, at this point, the campaign estimates savings of bandwidth is very limited. However, newer, low-cost of new technology is necessary andPractical technology (such as MPEG-4 and h.264 codec), to support the TV picture frame rate.

Realization of algorithm of compression standard MPEG video system is very popular.

MPEG actually contains more than one standard. MPEG-2 is an older standard that is mainly used for DVD and broadcast video, and you can target specific MPEG-4 bitrate more clear picture quality. H.264 is the latest increase of the standard, also known as MPEG-4 Part 10 or advanced video coding standard (AVC), which has a higher compression performance.

The general rule is that the standard 30 FPS analog video camera with H.264 compression technology, the bit rate is 1Mbps.

Usually MPEG-4 uses bandwidth than 50% of the bandwidth MPEG-2 than MPEG-4 50% more. To achieve this high-quality compression, running compression algorithm computing capacity will be increased accordingly, and this can lead to cost increases and delays for longer. However, as the signal handler function become increasingly powerful, prices are reduced, this calculation performance improvement of negative impact over the next few years.

System integrators or security managers must determine system acceptable minimum video quality, this determines the video image resolution and frame rate combined with the use of the codec, which determines the video data transfer rates, calculate the size and the amount of storage space.

At present, the algorithm used for the original system MPEG-2, while the new system is not using this algorithm. MPEG-4 is in an intermediate level of compression algorithm, used in the low-end systems.

[1] [2]