Industry News
    Home > News
  • The direction of development and its application of the intelligent network cameras

        Intelligent video analysis technology by high-speed analysis of massive image data, filter out useless information, users do not care about, but only provide useful critical information, and issue alerts and provide details, which can greatly enhance the operating efficiency of the video surveillance system from visual interpretation of video monitoring for automatic interpretation, this can be said to be a qualitative leap of industry, intelligence analysis must be video surveillance in the main direction of the future development.
        Intelligent video surveillance is on the rise
        Intelligent monitoring in the past few years has not been able to realize the the universality commercial, reason is nothing more than the poor quality of the video image can not provide basic resources for intelligence analysis, but this is two years with the increasing popularity of high-definition technology, so based on high-definition image intelligent analysis possible and the technology continues to mature.
        Intelligent Network Camera is the main form of front-end intelligent
        The front-end intelligence is one of the main application form of intelligent video analysis technology, network cameras and analog cameras due to the hardware structure of innate factors, which is completely different in the realization of intelligent mode. We know a smart cameras including the camera lens image acquisition link, A / D image preprocessing aspects, intelligent analysis and calculation of link encoding compression link, network transmission links five parts. The network cameras with image acquisition and processing, coding compression, network transmission module, simply add an intelligent analysis calculation module can be composed of intelligent network cameras, video image analysis and processing by the intelligent IP cameras can be based on the need to pass to the management center , and even some less demanding stronger CPU performance network cameras can simply poured into the intelligence analysis software will be able to achieve a series of intelligent applications.
        Analog cameras and its structure generally only image acquisition and processing aspects of image coding compression DVR or other external coding equipment to complete, built-in smart chip to achieve higher cost of smart applications directly on this structure, so the simulation the external smart box camera most of the front-end smart way. I.e. by Intelligence cartridge responsible for intelligent analysis of the image information provided by the camera, and then passes the results to the encoding device (such as a DVR, etc.), and then the encoding apparatus by this transmission to the management center. In other words, with the above integrated network intelligence different cameras, the intelligent analog cameras must pass multiple independent device operation to complete.
        Although smart application through an external intelligent box is also more common, but this split front-end intelligent solutions on the one hand and more equipment in the construction and maintenance of the more burdensome and high cost, on the other hand, multi-device interoperability run there will still be unstable and many other issues. By contrast, intelligent web cameras based on the DSP chip integration, regardless of to have a clear advantage in terms of cost of inputs or system safe and reliable, and scalable performance.
        Addition, and more importantly, the realization of the front-end smart premise is a high-quality image, the more perfect, the more abundant intelligent need the sharper image, before mentioned, intelligent applications to be truly universal must be universal access to high-definition monitor If you want to achieve HD and analog cameras must do not do compression processing of the video source for transmission, although the means of transmission of both analog YPbPr component transmission also digitized DVI, HDMI or HD-SDI transmission, but either of these cable and other means of transmission costs are very high, is impossible to truly universal application. HD IP camera is based on efficient compression, network transmission, the transmission cable is cable, IP transport with SD age exactly the same, and will not increase the cost of deployment, cost effective control The popularity of applications is certainly a very important factor. Shows in high definition, network HD cameras must be mainstream. And, as the network with high-definition technology continues to mature, the market size of the network HD cameras has expanded rapidly, models and performance are almost years with the development of analog cameras comparable, more developed, network cameras, integration The intelligent front-end applications will be more popular, intelligent video analysis boxes will slowly fade out of the market.
        Be focused application of intelligent network cameras
        Front-end smart with the Center Intelligence coexist
        Intelligent application front-end intelligent IP video camera relatively speaking, in the field of intelligent applications, there is another major form of is the central intelligence, also known as the back-end intelligent. Own chip calculated center smart way is to take advantage of the back-end computer software-analysis, that the camera does not provide any intelligent analysis of the result of the operation, and the use of front-end and front-end intelligent camera just to transfer image information management center, all all intelligent computing and analysis conducted by the Center.
        Center intelligence solutions in addition to these three modules, any analysis of the image due to the front of the camera does not constitute Centre intelligent image recognition, intelligent image recognition of the client and third-party data analysis in the Management Center To configure performance intelligent analysis computing software All image data must be transmitted to the management center, therefore, the center often also need one or more large capacity, superior performance and management server.
        Front-end smart Center Smart have their own strengths
        The front-end smart with the central intelligence caused differences in application performance due to their differences in the architecture program. First, the front-end smart three advantages in terms of performance, real-time high, and the camera to capture the video source immediately analysis generated immediately and report the results, which can effectively improve the operating efficiency of the system; camera video image analysis of identification can only be abnormal transmission of images to the center to record and display these images are generally up to account for only a few one-tenth of the total image data, so that you can greatly reduce the pressure on network bandwidth, to ensure the efficiency of the entire system running; tend to lose a part of the real information in video compression and produce some of the noise signal, the system if it is compressed video analysis and recognition process, will be prone to incomplete or inaccurate, and in the process on the front-end smart general analysis are collected uncompressed raw video processing and computing, so that effectively put an end to these problems. But the front-end smart Similarly facing some technical defects, the most significant is the large servers and computer software-based processing power compared to the performance and capacity of the front-end DSP chip, and the back-end is indeed very poor, often applied to the high in some intelligent analysis end of the industry, such as public security, due to the complex and diverse, and thus the amount of video analysis computation is very large, and most DSP but does not have a large enough memory and processing power to support smart performance requirements, low processing power will inevitably lead to analysis accuracy is low, resulting in the increase in false positives. Second, the scalability and flexibility of the front-end smart poor system when they need to increase or decrease the number of smart cameras or adjust the position, the operation will be a lot of trouble.
        Instead, the center intelligence by relying on strong server processing power to be able to achieve a very sound and excellent performance intelligent analysis, and the front-end resource scheduling and management is also extremely convenient. But on the other hand, as mentioned above, rely entirely on its center, the system must be equipped with a very well established network and server resources in order to cope with the vast amounts of image data transmission and management building costs, which will bring great challenges . In addition, in terms of time and accuracy, the mode of central intelligence showing certain drawbacks.
        Applications have focused on the front-end smart with the central intelligence
        First, the intelligent analysis of the actual use of the current is still the high-end industry-based front-end intelligent and center-based intelligent qualities in terms of performance, we believe that the former is more applicable to the specific behavior analysis places the center smart is more suitable for the synthesis behavior and historical data analysis, that the current traffic bayonet monitoring, Safe City, the practical application of the regulated industries have demonstrated significant.
        Traffic bayonet the field of smart applications tend to front-end.
        Prevention to reduce the occurrence of violations of road traffic violations, traffic accidents, and robbery of a motor vehicle traffic the bayonet surveillance system construction purpose, core performance requirements capture mainly including vehicles, license plate recognition, passengers face recognition. These properties must rely on intelligent video analysis can be achieved, and this is also the current intelligence analysis applications relative to the higher-end demand. The view from intelligent application needs in this area, such real-time requirements are more specific and are thus fully capable of smart cameras to complete.
        Safe city, regulated industries more suitable for central intelligence
        Safe City and regulated industries, although in a different form of expression of the intelligent monitoring application needs, but it can be classified as a large class. For example, the Safe City first need for complex monitoring sites: roads, squares, places of entertainment; Second, performance requirements, intelligent monitoring of the Safe City staff / goods stranded, abnormal behavior, illegal gathering, perimeter intrusion ; addition, safe city due to the very large number of video resources, which will inevitably be in the area of ​​real-time monitoring lacking rapid analysis, call intelligent application of historical data will be more urgent. Thus, intelligent monitoring of the Safe City from either place performance needs show a very complex, comprehensive, traceability, which correspondingly higher requirements for carrying smart computing platform, but smart camera chip processing power, or more reluctant. And the complexity of the monitoring sites also determines the system in the front-end link must have a strong flexibility and scalability, which is the previously mentioned disadvantage where the front-end smart.
        Prisons, detention centers, etc. The same is true monitoring system for regulated industries there is a complex place, but illegally gathered, abnormal behavior, illegal residence and other diverse needs of the same show a series of complex, integrated qualities Therefore, from the current view actual case, biased center smart solutions to obvious majority.
        Of course, both from the technical characteristics of the market demand characteristics for the front the intelligent and center smart both are not contradictory and even complement each other, complement each other, a video surveillance system is often not one or the other, and must be the front-end and the center smart and. With the continuous improvement of the processing capacity of the chip technology and the continuous improvement of the intelligent algorithms, intelligent internet camera market prospects have greater improvement even in the future, I believe that the two co-exist, have focused on the situation will remain so.