In the age of ubiquitous wisdom, the wisdom ecology is everywhere. As for the security industry, in the context of smart city construction, various market segments have an increasing demand for surveillance, and at the same time put forward higher requirements for large-scale networking and intelligent applications.
Huawei, switching to security is a late entrant, and is creating a new “end-cloud collaboration†AI security ecosystem based on AI cameras, striving to build an intelligent vision of the Internet of Things. When AI encounters security, the opportunity for latecomers to come from behind is here!
In the era of wisdom, traditional monitoring is hard to return
Security under the flourishing age has entered a period of major changes and major adjustments, and many shortcomings and deficiencies in the current video surveillance industry are also highlighted, which are reflected in the following aspects:
First, the more obvious the Matthew effect, the stronger the stronger and the weaker the weaker. It will be more difficult for production-oriented companies that rely too much on pure manufacturing and under-research and development. It is difficult to occupy too much market share and sustainable development in the low-profit game; some companies spend money on platforms or OEM products without core technology. Blindly following the trend, under this mixed status quo, increasingly rational market users will give feedback and screening in the shortest time to form the Matthew effect, the stronger the strength, and vice versa.
Second, in the face of complex scenes, traditional surveillance cameras are helpless. Especially in densely-flowed areas and some large-scale conferences and events, when there is a large flow of people, traditional surveillance cameras cannot guarantee the accuracy of capturing and extracting the objects that need to be monitored.
Third, the ability to take the initiative to prevent and intervene in the event is weak. For the public security and transportation departments, they most hope that surveillance cameras can meet the core demands of quickly discovering targets, locating targets, and tracking targets. At the same time, the current traditional surveillance cameras are mostly focused on post-event response, lacking proactive actions, proactive prevention, and even during the event. The ability to intervene.
Fourth, traditional surveillance cameras are becoming more saturated and technological updates are lagging behind, and the traditional surveillance solutions they build are difficult to meet the current user security and personalized business needs. For example, the single-point intelligent computing power is weak, the retrieval efficiency of the entire network is low, and the system hardware is complicated and difficult to operate and maintain. Another example: closed system data is segregated and difficult to share.
In the period of major industrial adjustments and major technological changes, for Huawei, an enterprise that has implanted technology into the bone marrow, the opportunity is really here!
AI empowers cameras, Huawei makes safe cities "smart"
From the above, it is not difficult to see that the security industry urgently needs a technological revolution to break the situation in the new era, which is to implant wisdom and make project construction "smart".
In Safe City, artificial intelligence and security have formed the best pair of CPs, which is a manifestation of the new kinetic energy contained in the new security. In order to make the safe city “smart†and provide more comprehensive and practical security and intelligent business guarantee, Huawei has launched an AI face capture camera at a time when artificial intelligence is popular. Road overtaking.
According to the reporter's understanding, Huawei's AI face capture camera has a built-in deep learning method and a face recognition algorithm trained through massive face big data, which can accurately locate and achieve precise capture in unconstrained scenes. Its millisecond-level positioning of key points of eyes, mouth, and nose contours can be adapted to various actual scenes such as large-angle side faces, blurred small faces, large expression changes, light and dark changes. At the same time, face detection and tracking can face the actual combat of dense crowd scenes, completely and effectively solve the shortcomings of the previous traditional cameras. It can be adapted to various actual scenes such as dense crowds, occlusion, side face, and blur. It realizes that the monitoring objects in the city can find the target object in the first time, and carry out the full face tracking, which meets the core demands of the safe city to quickly find the target, locate the target, and track the target.
Huawei products of this type have also made major breakthroughs in the core face capture function, surpassing the current human portrait capture machines on the market, and can detect and capture faces at full frame rate and full frame of video streams. The super performance of full frame rate and full-frame capture of portraits minimizes the probability of missed recognition. It is understood that the capture rate of this type of product is not less than 95%, supports low light 20lux, backlight 120dB, large angle 60 degrees, small face 35 pixels, face capture in unconstrained scenes; at the same time, it can effectively filter out duplicate faces. The repetitive face capture rate is controlled within 10%, effectively saving back-end storage and computing resources.
It can be said that Huawei's cameras after empowering AI will make the safe city "smart", no more fish will slip through the net, and all crimes and hidden dangers are under control.
Huawei builds a new “end-cloud collaboration†AI ecosystem based on front-end AI
The period of major adjustments and major changes in the security industry has also brought about a major trend of industry integration. Based on the AI ​​face capture camera, Huawei is building a new “end-cloud collaboration†AI ecosystem based on AI front-end applications.
There are not a few companies engaged in the research and development of intelligent monitoring in the industry, but most of them adopt the central intelligence model. The front-end uses traditional surveillance cameras to collect, and moves the processing and calculation of intelligent video analysis or face recognition to the back-end data center. Although this intelligent mode of the all-center mode can provide computing and storage support, it is too inefficient and will Bringing along with the rapid increase in the amount of data, as well as the pressure and cost issues brought about by the network transmission bandwidth.
In the future, artificial intelligence must be an end-to-end system, which can solve computing and storage problems in the front-end. Therefore, Huawei believes that for large-scale intelligent deployment scenarios, the "device-cloud collaboration" model of artificial intelligence is the most efficient.
When surveillance cameras are transitioning from HD to AI, Huawei’s “end-cloud collaboration†is to deploy AI face capture cameras at the front end to collect video and face image data, and complete the capture and face detection and tracking at the front end, while the back end cloud The platform performs face recognition and feature extraction and comparison of pictures, and realizes multi-scenario monitoring business applications with network deployment control, dynamic detection, and precise tracking. In this way, the pain points of traditional monitoring mentioned earlier are all solved.
For example, for the problems of weak single-point intelligence and low intelligence efficiency, "end-cloud collaboration" has extremely high intelligence efficiency and is fully capable of large-scale complex scenarios, while traditional monitoring methods are powerless in the face of complex scenarios; before It also mentioned that the operation and maintenance of the public security system and the use of resources are low. The "end-cloud collaboration" based on AI cameras can centralize discovery, centralized warning, report platform and unified management of resources; in terms of active prevention, end-cloud collaboration can be used throughout the network. Intelligence, analysis and response more quickly; for a series of data security issues such as video failure and loss, Huawei’s "device-cloud collaboration" adopts a full-scale four-fold guarantee of network layer, system layer, business application layer and business management layer, and security performance No. 1 in the industry.
The AI-based camera-based "end-cloud collaboration" model has many other aspects to crack the shortcomings of traditional surveillance methods, not to mention them one by one. In short, the new security has not only new kinetic energy, but also a new ecology. Based on front-end AI, Huawei is building a new AI ecology of “end-cloud collaborationâ€.
Concluding remarks
In addition to "end-cloud collaboration" to make security more efficient, at the Huawei Eco Partner Conference 2018, Huawei Security announced the concept of "Software-defined Camera". The first is to define the scene on demand-the future. Cameras should load different software and algorithms on demand according to different scenarios, and quickly adapt to the ever-changing pace of evolution through various combinations. The second is layered intelligence. On the one hand, smart cameras use precise judgments (from a single feature to full features) to improve accuracy and reduce misjudgments in a full-featured and multi-information manner. At the same time, through collaboration between multiple cameras , Collaboration between devices and clouds to make joint judgments, improve analysis accuracy, and support actual combat. Finally, there is continuous vitality-the camera should have the ability to continuously iterate, evolve, and grow algorithms and applications, and always meet the needs of customers' business scenarios.
It is undeniable that in security, Huawei is a latecomer, but for a company like Huawei that has always won with technology and is good at making trends, it is not uncommon for a company like Huawei to counterattack and overtake in other industries to become a leader. When artificial intelligence meets security, when the camera can be defined by software, it is a new opportunity for Huawei to make a big leap and make a big step forward. He will definitely follow the trend and strive to build a new ecosystem based on front-end AI capabilities.
So, under the new ecology based on front-end AI capabilities, what about security?
5V 3A Usb Charger,5V3A Wall Charger,5V3A Usb Wall Charger,5V 3A Usb Output Charger
Guangdong Mingxin Power Technologies Co.,Ltd. , https://www.mxpowersupply.com