Security surveillance cameras leap from analog to intelligent

From the development of the first generation of cameras to the present, the cameras have made great progress, from black and white to color, from ordinary guns to all-in-ones, from analog to digital, from SD to HD, and from non-intelligent to intelligent.

The camera is mainly composed of a lens, an image sensor, a DSP, etc. The subject is focused on the sensor through the lens. The sensor is composed of multiple XY pixels arranged vertically and horizontally. Each pixel is composed of a photodiode and related circuits. The light is converted into electric charges, and the total amount of the collected electric charges is proportional to the intensity of the light. The accumulated electric charges are moved out point by point under the control of the relevant circuit, filtered, amplified, and processed by the DSP to form a video signal output.

Analog age

In the age of analog surveillance systems. In the 1960s, black-and-white televisions and black-and-white cameras were gradually applied in the military field from the birth. As black-and-white cameras gradually entered the civilian field, an early video surveillance system was born, called a closed-circuit television monitoring system (CCTV). Analog transmission, also called analog monitoring system. At that time, the analog camera output was an analog video signal. The analog video signal generated by the video acquisition device can be converted into a digital signal by an encoder, and then stored in a computer. The video signal captured by the analog camera must be converted into a digital mode by a specific video capture card and compressed before being converted to a computer for use.

In the situation that high-definition, especially CMOS high-definition camera technology is still not mature enough, and the cost cannot be quickly reduced, analog cameras occupy the vast majority of the market share, and analog video surveillance has existed for a long time. Although high-definition cameras have been applied first in markets such as government and banks, for small and medium-sized surveillance systems such as residential quarters, shops, small supermarkets, and office buildings, analog cameras were still a cost-effective choice at the time.

Internet Age

With the advent of network video surveillance systems based on embedded technology, network and communication technology, and intelligent image analysis, the development of network video surveillance has entered the digital era. Whether it is a solution consisting of an analog camera plus a video server, a solution consisting of a network camera alone, or a solution consisting of a mixture of the two, IP surveillance has proven to be an attractive solution.

Into the network era, IP network cameras have become popular. In addition to all the image capture functions of general traditional cameras, network cameras also have built-in digital compression controllers and WEB-based operating systems. LAN, Internet or wireless network to end users. Network cameras can be directly connected to the digital network of TCP / IP, so the main function of this system is to perform video and audio transmission on the Internet through the Internet or internal LAN.

As the core equipment in network video surveillance systems, the application and popularity of network cameras represent the stage and degree of network video surveillance development. The network camera, known as the "Clairvoyance", has adapted to the current and future diversified needs and has become the mainstream product in the current video surveillance field. At the same time, the network camera products have more functional forms than traditional products, which can bring more added value to users. Camera manufacturers also incorporate rich and practical functions in their IPC products to attract more users.

Network surveillance cameras represent the future of surveillance cameras. The new generation of technology applications represented by high definition and ultra high definition is opening up new space for camera development.

HD era

In the SD era in the past, video surveillance was usually deployed in small environments with a single background and a small number of people. Even urban security surveillance cameras, due to the impact of their resolution, recorded video images that were blurred. Therefore, "seeing clearly" has always been one of the basic demands in the field of security monitoring.

In recent years, with the large-scale advancement of safe city projects across the country, the technology of security products, especially the technology of monitoring products, has developed by leaps and bounds. Among them, the core image sensor CMOS or CCD of the camera can meet the needs of millions of high-definition cameras. In addition, in the network camera, the high-definition products are not limited to the video processing technology of the analog camera, but also include the image enhancement technology, such as Image noise reduction, debounce, etc. Mastery and breakthroughs in this type of technology often require deep accumulation in digital video processing technology research and development and algorithms.

As the core area of ​​the current security industry, video surveillance is undoubtedly a key feature. The clearer the image, the more obvious the details, the better the viewing experience, and the higher the accuracy of applications such as intelligent analysis. Especially in the public security industry, through various previous cases and cases, it can be found that due to the low image quality, various major cases cannot be resolved in a timely manner, and the efficiency of detection is greatly reduced, so image clarity is the eternal pursuit of video surveillance.

Obviously, the value of high-definition images can only be realized when observing local images and wanting to obtain image details. For example, when a partial image is displayed in full screen, the image still has a very high resolution, and there is no mosaic phenomenon; when the electronic police or the road bayonet system recognizes a face, a number plate or a vehicle, a sufficient number of pixels can get a high Recognition rate. Especially with the development of artificial intelligence, the quality requirements for video images are getting higher and higher.

At present, the widely used 1080P Full HD has 2 million pixels and the resolution of 4K is equivalent to 4 1080P and 8 720P. That is, a 4K surveillance camera can be used as 4 quasi-HD surveillance cameras. Use ultra-high resolution to capture ultra-high-definition images for video surveillance, even if the picture is enlarged, it will not cause the picture to be blurred. 4K cameras can solve the problem of monitoring large scenes. Using a 4K camera can cover the square in front of the train station, and at the same time, it can basically see the appearance of pedestrians. Second, the same field of view, see more image details than 1080P cameras, such as human faces, license plates, etc., these details are often very critical in practical applications.

Enter the intelligent age

The camera introduced two or three intelligent analysis functions such as motion detection and video occlusion in 2011, and has developed to the present day. With the increase of computing power, the front-end camera can analyze video content, detect moving objects, and identify people in real time through the built-in artificial intelligence chip , Vehicle attribute information, and passed to the central artificial intelligence back-end database for storage via the network. At present, face identification, vehicle recognition applications, video structuring, and behavior analysis are common.

The current intelligent upgrade of the industry is not only reflected in the blessing of camera products AI, but also the implementation of scenes with practical and intelligent solutions.

At present, Hikvision defines AI applications based on scene requirements and launches AI multi-camera series products. Through the combination of multiple cameras and one device, it takes into account the needs of different perspectives, different parameters, and different functions in the same scene to improve the performance of single-point devices Data value density, to realize the perception of the entire scene, the extraction of all elements, and the association of all data; Dahua Co., Ltd. launched a new series of Ruijie based on its own chip products, including scene definition smart cameras, panoramic smart cameras, smart cameras, radar series, etc ; And Ruisi series products, such as intelligent video analysis box, edge intelligent analysis all-in-one, intelligent video compression server, intelligent network video storage server, etc., provide flexible edge computing power and lightweight industry applications of view intelligent services.

Whether it is an artificial intelligence startup company looking for a landing scene with the Huai algorithm, or a security manufacturing company based on traditional security business matching AI algorithms, let alone raising the digital banner of ICT giants, smart cameras are used as chips to promote the application of AI landings. Efforts to show the spark from the integration of security and artificial intelligence in scene-based applications. Of course, the intelligent upgrade of the industry is not only reflected in the AI ​​blessing of cameras and other products, but also needs to be combined with actual combat solutions as the propeller for scene landing.