Analysis of Video Surveillance Technology Points in Logistics and Warehousing Industry

In the logistics industry, customers reflect many problems and are roughly categorized. Basically, there are three types: poor nighttime lighting, poor visibility of indoor lighting, and impractical intelligence analysis. As far as the current video image technology is concerned, in order to solve such problems encountered in actual operations, core technology breakthroughs should be achieved in the three directions of wide dynamics, starlight level, and intelligent analysis, so as to achieve “being both clear and understandable. Effect". Let us now analyze how to proceed from these three directions!

The first, wide dynamic solution to the camera's backlight compensation problem

Wide dynamics mainly solves the problem of backlight compensation for cameras. In the loading and unloading area of ​​logistics and warehousing, whether or not the storage and handling of goods in freight cars can be clearly seen as a crucial and necessary factor, therefore, to improve the performance of cameras, how to break through the wide dynamic technology and make it mature, has become a measure of the performance of a camera. The important indicator. At present, the standard wide dynamic function has become the consensus of all IPC vendors. From the technical realization, wide dynamics are mainly divided into CCD+DSP and CMOS+DPS.

At present, most common surveillance network cameras use CMOS+DPS wide dynamic technology, the market demand is very strong, HD cameras generally have this function, so CMOS+DPS application scenarios are more extensive. Judging from the chip manufacturer's planning, the wide dynamic of multi-frame integration is a development trend, with smoother, more transparent and more dynamic range.

What is CMOS+DPS?

DPS uses a separate exposure and control technique for each pixel, plus a linear superposition of a full image synthesized from multiple frames captured using CMOS sensors, resulting in a higher dynamic range compared to double-exposure imaging of CCDs. Numerically speaking, a CMOS camera employing DPS technology can theoretically achieve a dynamic range of 120 dB. While expanding the dynamic range, DPS also solved the deficiencies of the CCD sensor in dealing with the dynamic range and color authenticity, and its color reproduction is more realistic and can meet the requirements of the application. Cameras using this technique use an analog-to-digital converter (ADC) in each pixel of the digital image sensor. When an optical signal is captured, it is directly converted into a digital signal, which minimizes the attenuation of the signal in the array. And interference.

Second, the appearance of the star-class low-luminance products completely solved night-monitored night dim images or infrared black-and-white images.

For video surveillance applications, low-light performance is often used as one of the reference standards for product selection. High-definition camera with excellent low-light performance can meet the clear color image under low light, and provide clear black-and-white images in all-black environments or under extremely low illumination.

Many manufacturers on the market today claim to have their own low-light star products, but many of them have the concept of stealing concepts. The true meaning of the high-definition starlight should be a smooth image with a resolution of not less than 1080P with a resolution of not less than 20,000 lux. To achieve this goal, we mainly look at two aspects of technology. The first is the choice of sensors. As is well known, in an image sensor having the same photosensitive area, the higher the pixel, the smaller the photosensitive area allocated to each pixel, the smaller the amount of light transmitted in one exposure period, and the lower the effect of low illumination. For example, all of them are 1/3-inch sensors, ordinary analog cameras are about 300,000 pixels, while high-definition cameras are at least 2 million pixels, and the pixels are 8 times that of analog cameras. The sensing area of ​​each pixel is only analog cameras. 1/8, if you do not use high-tech sensors, its high-definition quality at night can be appalling. At present, sensors with good low-light effects are mainly concentrated in some sensor manufacturers such as Sony and Aptina. Followed by the ISP adjustment, often can see different manufacturers of products using the same program, the same sensor, but the low-light effects are very different. This is because the ISP's noise reduction and gamma curve adjustment, the noise will make the image appear unclear, but the improper processing of noise reduction will result in serious fog, smear, loss of details and so on. If the gamma curve value is set too high, the contrast of the picture will be increased, the picture will be darker and the picture quality will be transparent, but the details will be lost. If the setting is too low, the picture will become brighter and the detail will increase, but the picture quality will become poorer. . This leads to the fact that even if they are based on the same hardware platform, the final results of the products will vary, depending on the manufacturer's own R&D capabilities.

Of course, any low-light camera needs a certain amount of light to use, and the degree of illumination involves the sensitivity of a low-light camera. So there is no 0LUX camera. It needs infrared light to make up for the camera to work. However, in some places, infrared cameras are banned from use, so so far, low-light cameras still have an irreplaceable position. If some important warehouses prohibit infrared rays, these sites must be monitored 24 hours a day. In some flammable and explosive warehouses, infrared cameras are also banned in these places due to the large heat generated by the infrared cameras, which can easily lead to spontaneous combustion and fire. Driven by the drive to monitor the Internet and market demand, Starlight-class HD cameras are bound to find a lot of applications because of their advantages in low-light and low-light applications.

The appearance of the star-class low-luminance products completely solves night-monitored night dim images or infrared black-and-white images. And as manufacturers complement the product line, Starlight products are pushed to almost all common application scenarios. Tiandi Weiye has launched a full range of products such as star light guns, star hemispheres, star ball machines, and star laser night vision equipment for various scenarios and product requirements. Starlight hemisphere is mainly used for indoor fixed-point monitoring, focusing on panoramic and large scenes; starlight bolting machine is mainly used for doorway fixed-point monitoring, focusing on depth scenes; and starlighting dome is mainly used for moving-point monitoring, focusing on close-up and long-distance monitor.

Third, intelligent analysis technology supports various smart applications and extensions, enabling regional alertness, population density, image anomaly, vehicle personnel flow statistics, vehicle comparison, face monitoring, and even facial recognition comparisons, through a combination of logistics warehouse monitoring Demand, provide data analysis for prevention and control work.

Intelligent analysis technology can be applied in many ways. Not only can it be integrated in the front-end camera, but also implemented in the back-end management platform software. It can be flexibly configured according to the actual project requirements and cost budget. The main advantage of intelligent analysis technology is to consider the accuracy rate. This puts forward specific requirements for the selection and deployment of on-site cameras. For example, according to specific scenarios (applied to gates, corridors, elevators, and even outdoor channels, etc.), consider the camera and the ground clip. Angle, angle between camera and target, lighting conditions, and weather conditions (cloudy sky, rain, and snow all have different effects).

Storage alert areas (access points, important areas, and fences) must be accurately monitored in all weather conditions in complex weather conditions (such as rain, snow, fog, and strong winds). When an anomaly occurs, the system will issue an alarm in real time, track the target trajectory, and record automatically. Post-event video tracing turned into effective prevention and prevention, and improved the safety factor of management. Although the application of video surveillance systems has become increasingly popular, most of them are viewing live video through manual monitoring. Because human attention is difficult to concentrate for a long time, the effect of manual monitoring is rapidly attenuated with the increase of events. To solve this kind of problem, intelligent video application technology came into being.

Intelligent analysis technology supports various intelligent applications and extensions, and can achieve regional alertness, population density, image anomaly, vehicle traffic statistics, vehicle comparison, face detection, facial recognition, and even facial recognition comparison capabilities. By combining logistics warehouse monitoring requirements, Prevention and control work provides data analysis. The main algorithm applications include: Area Alert - Automatic monitoring and alerting of personnel movement within the user-defined area, such as detecting whether a pedestrian or vehicle enters a predetermined area (or cross line) and prompting an alert. Behavior Analysis - Automatic monitoring and alarming of people fighting or strenuous exercise. Goods left behind/lost items--Automatic monitoring and alarm when the person or vehicle in the rest suddenly disappears when the suspect or the vehicle stays in the designated area for a long time or stays in the automatic monitoring and alarm. Face Detection - Face detection and tracking of pedestrians in a set area, screening of face images in real time and control, and provide a basis for later queries.

The above outlines technical breakthroughs in three directions for everyone, and hopes to further meticulously research and develop the logistics and storage security monitoring system, and can propose more effective and comprehensive solutions and technical guidance.