Intelligent video surveillance system design strategy (1)

As broadband wired and wireless network infrastructure improvement and global security market demand growth, video monitoring of the application as is explosive growth.

Video monitoring system development trend is clear, the digitization and networking, the next important trend that is intelligent, i.e. smart monitoring and video analysis technology.

Traditional video surveillance conducted by the manual surveillance video security breaches or abnormal condition or for the ex-post analysis, this application has its inherent weaknesses, it is difficult to achieve real-time security monitoring and test management.

With intelligent analysis function of monitoring system can distinguish between the monitoring object shape, movement, and other characteristics, proactive data collection, analysis, and based on preset criteria for alarm, recording, analysis and action. Intelligent monitoring system can run on the server, or you can run in embedded DSP-based system, which has gradually become mainstream.

Intelligent video applications generally can be divided into security, human behavior detection and intelligent transportation in three areas of application.

Which security application is widely regarded as the most potential market, and it includes several application categories: intrusion detection, you can automatically detect the video images of sport behavior; articles removed testing, can automatically detect the goods moving event — when a specific location within the control of goods being taken away or removed when the beep; remnants detection that can left of automatic detection — when items in a control zone is placed or abandoned when automatic warning; intelligent tracking, you can make the camera on its head and the zoom lens for autonomous PTZ drivers. Human behavior detection applications include tuogang detection (can automatically detect the sentry personnel in place), wandering detection (wandering on important regional human detection). Intelligent transportation applications include: illegal stay of transport to be detected, and when transport in preventing illegal stay in the region give alarm; vehicle retrograde detection, timely identification of vehicle retrograde.

As the accuracy and reliability to gradually increase and the product cost reduction, intelligent video in a growing number of occasions has been applied, it is able to replace part of the security equipment that reduce security staff strength, improve productivity, reduce administrative costs.

In fact, intelligent video applications with a very huge potential. As the technology matures, intelligent video technology applications is rapidly expanding, which includes the security, transport as well as retail and service industries, such as demographics, face recognition, crowd control, attention control and traffic control, etc.

Live video monitoring of the demand is growing fast, especially with real-time security monitoring application of the increased demand, real-time found security flaws or target exception behavior of features already have become increasingly important for relevance, intelligent video surveillance system products in this growing demand, is becoming a new video surveillance applications.

In particular, with the advances in semiconductor technology, such as the Blackfin converge-processor as the representative of advanced embedded solutions platform programme launched, has the extremely high cost performance and high availability of intelligent video analysis equipment has introduced and, in some key applications play an extremely important role.

Intelligent video applications design strategy selection of hardware platforms often determines the system's overall program cost, performance, development tools and methods available, as well as the programme and the feasibility of future continuous upgrade, so the programme selection is critical. platform

Intelligent video applications on its own unique demands on the hardware platform of choice for comprehensive trade-offs. Video surveillance system for networking and intelligent analysis of requirements, as well as large-scale projects installed on costs, volume and power consumption constraints, not standardized intelligent video analysis method and almost customised plan optimization mode, allows a combination of MCU and DSP advantage, with software design flexibility and powerful processing capacity of aggregation type processor programme platform reflects the more obvious advantage. This article will be combined with the analog devices Blackfin converge-unique DSP processor features, analysis of intelligent video design major design techniques.

First, the hardware platform for customizable capabilities is very important.

There are many factors to video surveillance system intelligent application steps: first, the intelligent monitoring video algorithm is more complex, difficult to standardize the various systems provider of video analysis software has its own unique algorithm, resulting in a product on the market, there is no uniform standards; second, video surveillance system more complex scenario, the user's requirements, so customization requirements. Therefore, video analysis programmes often need specific customer's application characteristics and needs of optimization, using the algorithm. In addition, due to the intelligent video applications of high complexity, processing capabilities to programmes put forward higher requirements. ASIC MCU + video monitoring traditional programmes difficult to achieve a variety of personalized design and high computing capacity requirements, even if you choose General DSP + dual-chip MCU programmes usually difficult to meet intelligent video monitoring application of complex operational requirements, the need for additional co-processor, this complex solutions regardless of cost, power consumption is the BOM or insufficient exploitation more difficult. Blackfin processor and give full play to the MCU + DSP together-the advantages of the schema to meet the intelligent video applications for system control and operation of high-strength requirements, in particular to BF561 represented by high-performance dual-core architecture has become intelligent video applications preferred platform.

Programme of scalability are factors to be considered.

Intelligent video analysis applications in addition to the need for application environment, application purpose optimization, different customers may also have different needs in other areas. For example, the current number of leading digital video surveillance programme enables H.264 base class @ Level3.0 and MPEG4 D1 + CIF dual stream support, the future may be extended to support H.264 D1 + CIF in dual stream. With the intelligent video analysis of wider application, such as IP cameras, wireless video surveillance, intelligent transport systems, different applications are possible for a variety of interface functions, communication standards, user interface, greater demand forDifferentiation, hardware platform on various requirements for flexible scalability is important. At the same time, as noted earlier, intelligent video analysis technology development but few years, as technology continues to mature, and some related standards promulgated and improvement, the product can be upgraded features essential both developer shall concern is also an important concern to end customers. Blackfin DSP algorithms for parallel processing with unique advantages, particularly ADSP-BF561 dual DSP core, can be realized very complex intelligent video processing algorithms.

Video application optimization characteristics.

Some programmes despite strong processing capacity and scalability, the programme is primarily for video applications optimized design also noteworthy because it is directly related to the design engineer available hardware design resources and system design difficulty and achievable performance. For example, Blackfin processors Blackfin for high strength, high data rate digital and media processing done a special optimization: Blackfin dozens of DMA channels and can be flexibly configured Cache is a great way to meet a video surveillance system on large computation, high data throughput requirements; ADI specifically developed fully optimized audio and video codecs, and freely available to large customers; application for video Blackfin incorporates many hardware drivers, including WiFi drivers, audio/video codec drivers; four video Blackfin arithmetic unit and video pixel instruction set significantly accelerate video operation speed; in intelligent video analysis of some base operator, such as a histogram, median calculation, Sobel calculation, the expansion in morphology, etc can use MIN and MAX of Blackfin directives to eliminate conditional jumps, saving processor cycles. Moreover, Blackfin also supports 13 non-video data vector operations. Appropriate design of data structures, in isolation, threshold before background calculation and update, and other links you can use the features of the directive allows Blackfin intelligent video analysis algorithms faster. These itself very effective instruction, most directives can be performed concurrently Blackfin, makes the processing power and then doubled.

Low power and stability is important.

Considering the intelligent video surveillance equipment are usually 7 days a week, run 24 hours a day, stability and power consumption is relatively important. In low power Blackfin processor on a variety of energy-saving technology: based on a gated clock kernel design, you can follow the instruction to selectively cutting power of functional units; support for multiple CPU for the desired action rarely during power-off mode; Blackfin processors support a self‑contained dynamic power management circuits, you can use this circuit on the operating frequency and voltage for independent control, in order to meet ongoing algorithm performance Blackfin processor required; most are supplied core on-chip voltage regulator circuit and low-voltage 0.8V. The Blackfin converge-unique architecture, a 90nm process, such as the leading low power consumption and disposal. Due to high processing capacity, Blackfin-based platform system can reduce the number of main chip, rich functionality and interfaces can meet a variety of peripherals and extensions need and reduce the number of components, thus ensuring greater stability and reliability. Present at the same price point DSP Blackfin DSP in low-power characteristics and stability is the best. [1] [2]