"Intelligent front-end" as a network monitoring technology can not only be applied to the traditional security fields such as alarms on crossing lines and tracking of moving objects; it is also gaining more and more attention in areas such as the statistics of people, facial recognition, and other business intelligence analysis. The importance attached. As people have a deeper understanding of the advantages of smart front-end, it is expected that their future application areas and functions will be rapidly developed.
I. Definition of intelligent front-end Currently, the intelligent video function of the network video system is mainly realized by means of central analysis. That is, through the video management software of the back-end server, intelligent analysis of the video stream transmitted by the front-end camera is performed to realize the front-end video information. Intelligent processing. What we are talking about today is "intelligent pre-positioning" in terms of the handling of this centralized server.
With the development of network video surveillance, the trend of digitalization has become increasingly apparent. As a digital terminal, the camera can not only perform functions such as video capture and recording, but also can perform some relatively high-end video analysis and processing through the CPU embedded in the network camera. The so-called "smart front-end" means that the video analysis functions or other management functions and intelligent functions that need to be processed are embedded in the camera, that is, what we call "video analysis and intelligent processing using cameras." .
The peculiarity of the “Intelligent Front End†network camera is that it can be based on an open platform and deploy more intelligent analysis applications to the front-end network cameras to meet the user's personalized video surveillance needs, which also means the same Due to the deployment of different smart pre-applications, the network cameras can achieve differentiated functions.
Second, the development of intelligent front-end network cameras 1. Smart front-end application is the development trend After adopting the “Intelligent Front-end†application, users can perform corresponding analysis in the front-end, and only need to pass the analysis and processing results to the background. When using a network camera to monitor whether suspicious personnel enter the space through the access control, if background server analysis is used, the video needs to be transmitted to the background every moment to analyze whether someone enters or leaves the space; if the front camera has intelligent analysis function, although the camera The entrance is monitored at all times, but no data is transmitted to the background in the normal state. When only the front end analyzes the state of someone entering or leaving, the camera automatically transmits the corresponding video to the background. If there are only ten minutes or more in a day when someone enters or leaves the space, then the entire network will be idle at other times.
At present, the camera end can be embedded with a local SD card or other corresponding storage device. Users can even store video directly to the camera without the need for background network transmission, which can further reduce the burden on the network. Reduce video transmission costs. The direct analysis and processing by the front-end camera is less restricted by other factors, and the impact on security is relatively less.
2. The opening of chip resources is the key to development. Openness is the most significant feature of the network video surveillance era. With the improvement of the chip performance of the camera, more chip resources will be opened up. Some companies that specialize in video analytics for smart video can Based on open chip resources, the corresponding program is embedded in the camera to help achieve better front-end intelligence. At that time, not only the external interface of the camera is open, but also the network protocol involved in the camera is open, which can achieve the compatibility of many third-party video analysis programs and facilitate the user to make the best combination among many software and hardware manufacturers.
Intelligent front-end, on the other hand, can also be understood as an application platform on the network camera side. In the future, cameras will not only play a basic role in video acquisition and transmission, but also run third-party video analysis programs.
For example, the most popular iPhones currently running can run many APP applications, and there are also many program services that are tailored to a specific group of people. These program services make iPhones not only exist as a communication tool. The same goes for web cams. In the future, not only video will be transmitted, but also embedded in various software applications. Some customized smart analytics functions can be run on our cameras. Therefore, with the development of network cameras, the front-end smart development must be based on a more open application platform.
Third, the advantages of intelligent front-end Intelligent analysis is equivalent to give the video surveillance "brain", is the future development trend, and the smart front will also be in the development trend of intelligent video surveillance has been made great progress, and promote the entire smart The development of video surveillance. The author summarizes the advantages of intelligent front-end as follows:
1. Saving network bandwidth to process webcam video in the background requires the video to be transmitted to a back-end server, and then processed and analyzed according to the video content. In this case, it is easy to waste a lot of bandwidth on unnecessary video transmission, but also for system storage. Certain pressure.
The front-end video analysis function is put into the front-end processing for some analysis work. After discovering the situation, it is transmitted to the background to process the corresponding result. The entire system has a relatively balanced distribution of its resources or processing pressure and also saves a lot of network bandwidth. .
2. Increased scalability Each chip in the smart front camera and intelligent software allow the camera to run independently with independent video analysis capabilities. No matter how many cameras are in a system, cameras can be added at any suitable place to achieve a good extension function, and it will not increase the burden on the server for the background.
3. The more accurate false alarm rate belongs to the problem that needs to be overcome in the development of the entire network monitoring intelligence, and has nothing to do with intelligent pre- or post-facility. However, if the camera obtains video data, if it is transmitted through the network to the back-end, some network delays, packet loss, and compression may occur during this process, which will also affect the accuracy of the analysis. The front-end analysis is performed directly at the camera end, which can make the video processing result more accurate.
Although the current intelligent video analysis still does not achieve 100% accuracy, it can at least help customers solve some very practical problems. For example, the front-end can distinguish the alarm state from the normal situation video and help the customer to quickly retrieve the current situation. This selection of video analytics development technology is also moving toward higher speed and intelligence.
4. Saving costs Implementing intelligence in the back-end server requires many servers to perform analysis and processing, and a large amount of hardware and service costs are incurred in system configuration and maintenance. The opening of front-end camera resources to achieve intelligent front-end only moves the video processing functions forward, without adding any cost. On the contrary, a server can handle video analysis functions of 20 or 30 channels at most. If some analysis processing is moved to the front end, the server only needs to handle alarms, which will save some costs of the background server.
In addition, intelligent analysis software itself is more expensive. In the future, if more hardware and software vendors cooperate in a more open manner, it will inevitably bring more third-party video analysis software engineers to join this platform and upload more intelligent analysis functions to the cameras. The end performs more convenient operations and uploads through the flexible video analysis module, which reduces the overall cost of intelligent video analysis.
to sum up:
Smart front-end is to move the video analysis function to the front-end for processing. On the one hand, it can save background resources, let each camera handle the video analysis function independently, with better accuracy and better scalability. Upgrade capability.
The author believes that the future direction of development is the front-end to process video, the preliminary analysis results will be transmitted to the background, and the background according to the results to take appropriate alarm measures, these measures may be to activate a video or sound and light alarm, send e-mail, etc. Wait. The back end will not need to perform video processing or video analysis work, but only accept the corresponding processing results, and then take corresponding measures on the basis of this result, so the future is a development process in which the front end and the back end are combined with each other.
In addition, in the current trend of intelligent network monitoring has been unstoppable today, we need more resources are open, shared vendors, the development of camera-side CPU resources open, you can let a lot of third-party professional video analysis, video analysis software to write To the camera, it integrates the strengths and advantages of each company to work together, giving customers more room to choose.
I. Definition of intelligent front-end Currently, the intelligent video function of the network video system is mainly realized by means of central analysis. That is, through the video management software of the back-end server, intelligent analysis of the video stream transmitted by the front-end camera is performed to realize the front-end video information. Intelligent processing. What we are talking about today is "intelligent pre-positioning" in terms of the handling of this centralized server.
With the development of network video surveillance, the trend of digitalization has become increasingly apparent. As a digital terminal, the camera can not only perform functions such as video capture and recording, but also can perform some relatively high-end video analysis and processing through the CPU embedded in the network camera. The so-called "smart front-end" means that the video analysis functions or other management functions and intelligent functions that need to be processed are embedded in the camera, that is, what we call "video analysis and intelligent processing using cameras." .
The peculiarity of the “Intelligent Front End†network camera is that it can be based on an open platform and deploy more intelligent analysis applications to the front-end network cameras to meet the user's personalized video surveillance needs, which also means the same Due to the deployment of different smart pre-applications, the network cameras can achieve differentiated functions.
Second, the development of intelligent front-end network cameras 1. Smart front-end application is the development trend After adopting the “Intelligent Front-end†application, users can perform corresponding analysis in the front-end, and only need to pass the analysis and processing results to the background. When using a network camera to monitor whether suspicious personnel enter the space through the access control, if background server analysis is used, the video needs to be transmitted to the background every moment to analyze whether someone enters or leaves the space; if the front camera has intelligent analysis function, although the camera The entrance is monitored at all times, but no data is transmitted to the background in the normal state. When only the front end analyzes the state of someone entering or leaving, the camera automatically transmits the corresponding video to the background. If there are only ten minutes or more in a day when someone enters or leaves the space, then the entire network will be idle at other times.
At present, the camera end can be embedded with a local SD card or other corresponding storage device. Users can even store video directly to the camera without the need for background network transmission, which can further reduce the burden on the network. Reduce video transmission costs. The direct analysis and processing by the front-end camera is less restricted by other factors, and the impact on security is relatively less.
2. The opening of chip resources is the key to development. Openness is the most significant feature of the network video surveillance era. With the improvement of the chip performance of the camera, more chip resources will be opened up. Some companies that specialize in video analytics for smart video can Based on open chip resources, the corresponding program is embedded in the camera to help achieve better front-end intelligence. At that time, not only the external interface of the camera is open, but also the network protocol involved in the camera is open, which can achieve the compatibility of many third-party video analysis programs and facilitate the user to make the best combination among many software and hardware manufacturers.
Intelligent front-end, on the other hand, can also be understood as an application platform on the network camera side. In the future, cameras will not only play a basic role in video acquisition and transmission, but also run third-party video analysis programs.
For example, the most popular iPhones currently running can run many APP applications, and there are also many program services that are tailored to a specific group of people. These program services make iPhones not only exist as a communication tool. The same goes for web cams. In the future, not only video will be transmitted, but also embedded in various software applications. Some customized smart analytics functions can be run on our cameras. Therefore, with the development of network cameras, the front-end smart development must be based on a more open application platform.
Third, the advantages of intelligent front-end Intelligent analysis is equivalent to give the video surveillance "brain", is the future development trend, and the smart front will also be in the development trend of intelligent video surveillance has been made great progress, and promote the entire smart The development of video surveillance. The author summarizes the advantages of intelligent front-end as follows:
1. Saving network bandwidth to process webcam video in the background requires the video to be transmitted to a back-end server, and then processed and analyzed according to the video content. In this case, it is easy to waste a lot of bandwidth on unnecessary video transmission, but also for system storage. Certain pressure.
The front-end video analysis function is put into the front-end processing for some analysis work. After discovering the situation, it is transmitted to the background to process the corresponding result. The entire system has a relatively balanced distribution of its resources or processing pressure and also saves a lot of network bandwidth. .
2. Increased scalability Each chip in the smart front camera and intelligent software allow the camera to run independently with independent video analysis capabilities. No matter how many cameras are in a system, cameras can be added at any suitable place to achieve a good extension function, and it will not increase the burden on the server for the background.
3. The more accurate false alarm rate belongs to the problem that needs to be overcome in the development of the entire network monitoring intelligence, and has nothing to do with intelligent pre- or post-facility. However, if the camera obtains video data, if it is transmitted through the network to the back-end, some network delays, packet loss, and compression may occur during this process, which will also affect the accuracy of the analysis. The front-end analysis is performed directly at the camera end, which can make the video processing result more accurate.
Although the current intelligent video analysis still does not achieve 100% accuracy, it can at least help customers solve some very practical problems. For example, the front-end can distinguish the alarm state from the normal situation video and help the customer to quickly retrieve the current situation. This selection of video analytics development technology is also moving toward higher speed and intelligence.
4. Saving costs Implementing intelligence in the back-end server requires many servers to perform analysis and processing, and a large amount of hardware and service costs are incurred in system configuration and maintenance. The opening of front-end camera resources to achieve intelligent front-end only moves the video processing functions forward, without adding any cost. On the contrary, a server can handle video analysis functions of 20 or 30 channels at most. If some analysis processing is moved to the front end, the server only needs to handle alarms, which will save some costs of the background server.
In addition, intelligent analysis software itself is more expensive. In the future, if more hardware and software vendors cooperate in a more open manner, it will inevitably bring more third-party video analysis software engineers to join this platform and upload more intelligent analysis functions to the cameras. The end performs more convenient operations and uploads through the flexible video analysis module, which reduces the overall cost of intelligent video analysis.
to sum up:
Smart front-end is to move the video analysis function to the front-end for processing. On the one hand, it can save background resources, let each camera handle the video analysis function independently, with better accuracy and better scalability. Upgrade capability.
The author believes that the future direction of development is the front-end to process video, the preliminary analysis results will be transmitted to the background, and the background according to the results to take appropriate alarm measures, these measures may be to activate a video or sound and light alarm, send e-mail, etc. Wait. The back end will not need to perform video processing or video analysis work, but only accept the corresponding processing results, and then take corresponding measures on the basis of this result, so the future is a development process in which the front end and the back end are combined with each other.
In addition, in the current trend of intelligent network monitoring has been unstoppable today, we need more resources are open, shared vendors, the development of camera-side CPU resources open, you can let a lot of third-party professional video analysis, video analysis software to write To the camera, it integrates the strengths and advantages of each company to work together, giving customers more room to choose.
KMP Parts,KMP Engine Parts,KMP Auto Parts,KMP Excavator Parts
JINING SHANTE SONGZHENG CONSTRUCTION MACHINERY CO.LTD , https://www.sdkomatsuexcavatorparts.com