To make intelligent network video surveillance system , the most important way is to make the network cameras intelligent. After analyzing surveillance video to identify the front-end intelligent network cameras, you can only have abnormal image transmission to the rear to record and display, which can greatly reduce the width on the network. But how to make the surplus CPU resource of network cameras, intelligent video analysis and recognition which the software embedded algorithm to the network camera, is currently one of the key
China network camera manufacturers.
As we all know, get a photoelectric image depends cameras Analog cameras can only transmit one-way video signal, its performance is subject to the technical capacity DVR and operator. But the network camera can transmit full-duplex digital information, and can be effectively integrated with other devices within the system, to enable them to achieve higher levels of performance in a distributed and scalable environment. Network cameras can communicate in parallel with multiple applications, in order to achieve a variety of functions, such as to detect the movement of the screen, or transmitting the video streams of different formats and the like. IP address of the camera with more convenient network, because it does not require additional analog / digital signal conversion, do not need to send the device in the camera-side installation, so the camera's size will become smaller and smaller, and this is easier subtle, security monitoring system is needed.
Network cameras make traditional cameras and computer functions into together, it is not only the video signal is digitized and compressed, also has a network interface. Its video transmission via IP network, and recorded on a standard PC with video management software. Systems consisting of a network camera is a true network video system, it is also a fully digital system, because there is no use of any analog device. Monitoring scenes video signals from a network camera continuously through the IP network, regardless of where the viewer is located, the camera can give the viewer to provide a stable image quality. Therefore, the network camera most suitable for application in such a large city surveillance network security monitoring systems, and to build into intelligent network security monitoring system, the first network camera should enable intelligent. Thus, the monitored video images, after analysis to identify the front-end intelligent network cameras, you can only have abnormal image transmission to the rear to record and display, which can greatly reduce the burden on the network. But as the intelligent software algorithms added all front-end network camera, due to the limited CPU resources of its own, often need to improve the front-end equipment configuration or use a dedicated video analysis equipment together, it would be the size and cost of the device is greatly increased, thus nor is the user required.
Function module HD IP network camera solution compared to conventional cameras, the camera's internal network is to increase the digital processing and digital video compression and network access, forthcoming digitized video signals into compliance with network transport protocols (such as TCP / IP protocol) data stream for transmission to the internet. Thus, a network camera can be viewed as a combination of cameras and a computer. It can capture images directly transmitted over an IP network, allowing authorized users through a standard IP-based network infrastructure, in the local or remote locations to achieve view, store and manage video data, or monitor the camera's built-in microphone collection site sound.
Since the network camera also communication (COM) interface that allows a network camera has to capture and conversion control data. Because through the COM interface can be connected to an external decoder. Therefore, the network took an authorized user of a PC, you can send commands transmitted over the network via the camera COM port to its external decoder, the camera head and lens necessary action. Meanwhile, the COM port can accept data from the front end, and transmitted to the network on either a PC.
The network camera also alarm (ALARMI / O) interface. This alarm signal can be received by the access terminal came various types of alarm sensor switch signal, and encodes, and then into the TCP / IP network, a variety of joint response settings (such as through the application of software for the alarm signal visual and audible alarm and alarm recording, etc.).
Because each network camera has its own IP address, it can be connected directly to the network, and built-in Web server, FTP server, FTP client, Email client, alarm management, programmability, and many other functions. Network cameras without connecting with the PC, it can be run independently, and can be placed in any location with IP network interface. In addition to the video information, the network camera can also go through the same network connection more other functions, and transfer some of the other useful information, such as video motion detection, audio, digital input and output (can be used to achieve the alarm linkage, such as trigger alarm or activate the on-site lighting, etc.), for transmission of serial data or a serial port PTZ device drivers and the like. Network camera image cache, you can save and send video images before and after the alarm.
Thus,
IP network camera video source can be obtained at local pretreatment, compression, and for remote transmission and storage over IP networks. An important part of containing video pre-processing, video compression, and IP network video (Video-OverIP) function.
Due to the current IP camera resolution is D1 gradually evolved from the standard definition to HD, and must be a local real-time compression, which can only use hardware compression mode. If you use multi-chip DSP processor, will make the system cost and increase integration and power consumption, while allowing users to not accept; if low-cost single-chip FPGA devices use, performance can not meet the design requirements, but if the use of single high-performance FPGA devices, we can meet the requirements. Because this device is structured with a corresponding series of devices, it can further reduce the cost to 1/10, and 50% lower power consumption.
Composition and implement the principles of intelligent network camera solutions
By foregoing the use of single-chip high-performance Stratix series FPGA devices function module program per HD IP network camera, with an embedded intelligent software algorithms module can be composed of intelligent network cameras. Obviously, six parts of a smart network cameras, that is, from the camera lens, CCD or CMOS imaging device, A / D and preprocessing, intelligent software algorithms module, coding compression module section with a network interface module.
The image pickup lens of the optical image of the scene being monitored to collect and focus them imaged onto a photosensitive surface of a CCD or CMOS imaging device, an electric signal converted video output after A / D conversion and noise removal, image enhancement and other pre-processing, sending embedded intelligent software algorithms into modules. Intelligent software algorithms are the most important part of the intelligent network cameras, the algorithm of the pros and cons, abnormal signal analysis identified whether accurate or not is the camera key practical. When the intelligent network cameras began to enter the monitoring state, the will of the encoded image compression and transmission to the scene to monitor the network control center, and then you only have to identify abnormal signal, only the encoded compressed, the interface to the transmission network through This can greatly save network bandwidth resources.
Because video compression and transmission by post, will lose part of real information and generate some noise signal, then use intelligent software for video compression after analysis and identification process, there will be relatively more leakage alarms and false alarms. The intelligent network camera of this program is no compression of the original image object extraction, it is clear that there is no such problem. And, it is only in the event of exceptional circumstances required under the pre / alarm, only need to be sent to the rear end of the associated video monitoring and recording, but generally only rarely transmitted through the network target data, the target data information The flow rate is less than 1/50 of video traffic, thereby reducing network bandwidth requirements and consumption, reducing the user's investment in the network, while at the same time on the network condition is not good to bring users to use intelligent video network possible monitoring system.
Therefore, the focus intelligent network camera program is a good choice and design of intelligent software algorithms, and then embed DSP or FPGA to form intelligent software module section.
Design principles and elements of intelligent network camera program
Design principles and elements of the intelligent network camera solutions are the following.
Modular design
Here, the modular design, the main function is to make the intelligent modularity. The author believes that speaking in front of that can be embedded in the intelligent network cameras 8 intelligent features, except for the first three basic functions of each camera is required which consists of a module, the rest can be composed separately. This is more flexible and convenient, you can monitor the site according to the actual needs and choose different modules loaded, thereby reducing capital investment, engineering companies are willing to accept. Thus, modular design can not only bring convenience to manufacturers, but also very suitable for practical needs.
Reliability Design
Any product should be subject to reliability design, which requires a reasonable system designed structure, product durability, the system is stable and reliable. Thus, the need for EMC design, to be anti-EMI, lightning and all kinds of interference in order to adapt to harsh climatic conditions without failure.
Humanized operation design
The so-called humanized operation design, which is very suitable for people of simple and convenient operation. As far as possible
Push-button operation, there must be a clear indicator and horn tips, it was hidden in its installation and high, can also use the remote control, will focus on the daily operation of the remote control.
Practical design
The so-called practical design, the system is very practical, easy to use, with fewer links, easy to learn. Can be easily set up automatic function, as well as theft, fire, access control systems linkage, high degree of automation, the system can occur in case of an actual infringement of timely pre / alarm. In addition, also requires a self-test function and self-test results regularly reported to the management center, accept the management center monitoring, in order to maintain failure immediately.
Economic Design
When considering the economic system, in addition to considering lower development cost of the system, outside the production costs, but also take into account the system's cost, including maintenance fees already during use, spare parts fees.
In order to obtain a higher price, the design should not blindly pursue complex advanced solutions. Under the premise of meeting the performance indicators, as far as possible using a simple program, because the program is simple, simply mean that a structure, high reliability and low cost.
After the above factors into account, then select OK appropriate programs.
