To say that a surveillance camera has good definition and high definition is a descriptive statement without quantification. As a user, I don’t quite understand how clear the camera is and what is good about it. At the same time, on many occasions, it is not convenient to compare with other similar cameras. The purpose of the parameters is to make the quality of the product, the function can be standardized, quantified, and can more accurately describe the corresponding security product performance.
The reality is that the original objective parameter data, there is artificial fraud, confusing the situation. Security camera parameters are randomly marked, and there are many cases of false marks, some of which are intentional, and some of which are not clear to the manufacturers themselves and are randomly marked. Here is a summary of the common parameters of surveillance cameras.
Minimum illumination
The lowest illumination refers to the low-light performance of the camera, that is, the imaging ability in the case of low light, fill light or no light. The illumination of the environment is expressed in illuminance (in lux, lux). 1lux is approximately equivalent to the brightness of a candle illuminating 1 cubic meter of space.
In addition to the ambient light conditions, the low-light performance of the camera also affects the aperture size of the lens, the photosensitive performance of the sensor, and the quality of isp image optimization. Each parameter is not expanded here. Therefore, the accurate expression of a camera’s low-light performance should be as follows:
color: 0.005 Lux @(F1.2,AGC ON),0 Lux with IR
source: Hikvision’s official website A certain product
This means that this camera uses an aperture F1.2 lens to turn on the automatic gain (AGC) in ISP image processing, and can maintain a color image under 0.005Lux lighting environment (below 0.005Lux, the image is black and white). The latter 0 Lux with IR means that the image can still be imaged normally in a completely dark environment (0 Lux) using infrared supplementary light.
Regarding the random labeling of the minimum illuminance parameter, the negative example will not be given. We clearly define several key factors that affect the minimum illumination of the camera: lens aperture size, sensor’s own low-light performance, ISP image processing, ambient illumination (Lux, whether there is a supplementary light, etc.).
Let’s look at another example. A certain camera of Dahua expresses its minimum illumination like this:
0.002Lux (color mode); 0.0002Lux (black and white mode); 0Lux (infrared light on)
source: dahua official website
The controversial point is 0.0002Lux (black and white mode), how to understand the black and white image under 0.0002Lux. The image goes from color to black and white, with a critical point in the middle, such as 0.002lux (color mode) in this example. That 0.0002Lux is the critical point of black and white images to what kind of images? Obviously, such a parameter expression is not accurate.
Possibly to avoid misunderstanding, Haikang’s parameters omit the minimum illumination in black and white mode. The Axis provides us with a more accurate statement:
Color: 0.16 lux at 50 IRE, F2.0 B /W: 0.03 lux at 50 IRE, F2.0 0 lux with IR illumination on
source: Axis
It introduces a parameter-IRE. Iire is a unit for measuring a video signal (level). The effective part of the video signal-video completely Black (Black level) to video completely White (White level) is divided into 100 parts, which is defined as 100IRE units, I .e. 0~100IRE(700mV).
A video of 100IRE(700mV) represents the video image with the best brightness and contrast, while a video of 50IRE(350mV) represents only half the contrast. Usually 30IRE(210mV) is the lowest number of available video images. (Axis chose 50IRE as the standard.)
A low illumination result measured at 10IRE can be 10 times higher than a result measured at 100IRE, so a result that does not mark the IRE level is actually meaningless.
Lens Aperture
There is not much to say about this. There are false and disorderly marks. The lens aperture affects the camera’s low-light performance and depth of field. The larger the aperture (the smaller the F value), the greater the amount of light, the shallower the depth of field.
The amount of light entering the F1.0 lens is twice that of the F1.4 lens, so it can naturally have better low-light performance.
The larger the aperture, the shallower the depth of field, which means that the monitoring screen can tolerate before and after the unclear (focus is not clear) the lower the degree. Therefore, in practical application, the larger the aperture, the better.
Resolution
Two issues are involved here, the resolution standard and the difference method to artificially improve the physical resolution of the sensor.
First of all, let’s talk about the resolution standard. Regarding the resolution, the standards in different fields are slightly different and are not uniform and clear. We should have a full understanding.
When we say a 2MP camera, we mean that the effective pixel of the camera is 2 million pixels. A pixel is a point. For a picture or a two-dimensional video screen, these pixels are distributed in both horizontal and vertical directions (horizontal resolution and vertical resolution). In General, 2MP cameras have 1920 pixels in the horizontal direction and 1080 pixels in the vertical direction, 1920 x 1080 is equal to 2073600, that is, more than 2.07 million, which is the origin of 2MP cameras. There are 3MP,4MP,5MP,8MP and so on. This is from the perspective of total effective pixels.
There is also a standard from the viewpoint of effective pixels in the vertical direction (vertical resolution). For example, 720p (1280 x 720),1080p (1920 x 1080). In the film and television industry, 720p is named HD (high definition) and 1080p is named Full HD (Full HD). Then it is Ultra HD (UHD), which generally refers to the resolution of 3840 x 2160 (8 million pixels, 4096 x 2160, also known as 2160p) or 4000 x 3000(12 million pixels), and some refer to 8K(7680 x 4320) as UHD.
The Internet industry from the perspective of horizontal resolution, and the rise of another naming rules. 1K (such as 1080p) because the horizontal resolution is close to or greater than 1000. 2K (such as 2304 x 12,963 million pixels),2.5K(2560 x 14,404 million pixels),3K(2880 x 1620 or 2592 x 19,445 million pixels),4K(3840 x 2160), and so on. According to this naming rule, we can call the currently popular two 2 million-pixel binocular stitching cameras (for example, two camera images 1920 x 1080 pixels are stitched left and right) 4K cameras (the stitched pixels are 3840 x 1080).
The second mess about resolution is the difference method of artificially increasing the resolution of the sensor. For example, we are familiar with a Sony sensor — IMX307 with a physical resolution of 2 million. Many camera manufacturers have improved the resolution of cameras using this sensor to 3 million through the difference method. The purpose is obvious.
The specific implementation method of the difference is not expanded here, mainly talking about the problems caused by it.
Although the picture resolution has been improved, the monitoring angle of view and the exquisite degree of the picture have not been improved. The lens with the same focal length is used on the camera with and without difference. The picture is exquisite and the angle of view is very different (generally, the size of the sensor target surface with low resolution is smaller, so the angle of view will be narrower), and the contrast is obvious. By comparing the monitoring perspective, it is easy to find out the difference scheme.
For price-sensitive consumer security cameras, the common practice is to change the 720p sensor to 1080p, while the 720p sensor target surface is mostly 1/4 ‘, compared with 1/3’ 1080p sensor, the monitoring viewing angle difference is very obvious.
Number of Wireless Antennas
Wi-Fi and 4G (and even 5G) have many applications in security products. Such as Wi-Fi cameras, Wi-Fi NVR,4G cameras and so on. One of the chaos in these products is to increase the number of antennas at will. The subtext is that the more antennas, the more stable the wireless signal of the product, the more advanced the technology, and of course the price should be more expensive.
Generally speaking, for ordinary Wi-Fi cameras (whether 2.4g or 5G band, or dual band), few support 2 x 2MIMO dual Wi-Fi technology. One is because the Wi-Fi module is required to support 2-way transmission and 2-way reception (2T2R), which significantly increases the cost. The purpose of the two-antenna is mainly to increase the upper limit of the physical bandwidth of wireless signal transmission. For a single camera with little bandwidth requirement, it is completely unnecessary. On the NVR, using 2X2MIMO dual antennas is a good choice (because of the simultaneous management of multiple Wi-Fi cameras).
An interesting phenomenon is that many Wi-Fi cameras on the market use 2 antennas, even 3 antennas and 5 cameras. Careful inspection will find that except for one normal Wi-Fi antenna, the rest are all fake (only the appearance of the antenna, there is no Wi-Fi signal line inside).
Another security product that requires a wireless antenna is a 4G network camera. One point that needs to be made clear is that if you are using the LTE cat.1 module, only one antenna is needed, and if you are using the cat.4 module, you need two antennas.
Zoom Magnification
Security cameras that support optical magnification can see more distant targets and are widely used in daily applications. The reality is that the lens that supports optical magnification is expensive, and the larger the magnification, the more expensive the lens is.
The magnification of the lens zoom is generally displayed on the monitor screen (OSD). Some manufacturers have used their brains on the lens magnification numbers displayed by the OSD. Originally 18X lens, changed to 30X on OSD display. To impersonate a 30-fold magnification camera. Change from small to large.
Although the zoom number superimposed by OSD has become larger, the actual multiple of the lens has not changed. It is easy to deceive the masses (users) who do not understand the truth.
We need to focus on the camera’s optical zoom, not the digital zoom.
IP Protection Level
IP protection level refers to the waterproof and dustproof ability of the camera shell, and there is a corresponding standard. For example, Chinese standard GB/T 4208-2017 (with European standard IEC 60529:2013). NEMA 250-2018 in the United States, JIS C 0920:2003 in Japan, UL50 standard for commercial certification, etc.
IP66, the first number indicates the ability of the camera shell to prevent dust. The larger the number, the stronger the dust-proof ability. The second number indicates the waterproof ability, the larger the number, the better the waterproof ability. Because there are standards, there are corresponding test requirements. However, some manufacturers have not been tested and randomly marked the IP protection level of the camera.
The corresponding IK explosion-proof rating.
WDR Value
As with the lowest illumination, the camera width dynamic value also has the condition of random mark. One is to confuse real wide dynamics (true WDR) with digital wide dynamics (dWDR,digital WDR). The other is the virtual WDR value, for example, the original only support 80dB WDR, virtual mark is 120dB.
Night Vision Distance
The general low-light performance of the camera is not so powerful. In the case of night or poor light, to obtain a good image effect, it is necessary to fill the light. Common fill lamps include infrared lamps (850nm,940nm), white light (warm light), etc.
Here involves a basic common sense, the illumination distance and angle of the fill light should be greater than the original camera lens (daytime visible light) visual distance and viewing angle.
Some manufacturers mark or exaggerate the camera infrared, white light supplement light irradiation distance. Of course, there is something understandable here. The visual angle of the surveillance camera is easy to see and measure (for example, 90 °,60 °, etc. are easier to measure and observe with naked eyes), but the visual distance is more difficult to determine. For example, how to judge a camera can see 80 meters or 100 meters distance, with what standard.
This involves another problem. How to divide the visual distance of the camera by pixel density. Different pixel densities correspond to different monitoring functions and requirements, such as Detection, Observation, Recognition, and Identification.
According to EU standards (EN62676-4:2015, EN50132-7:2012, detection requires 25 pixels/m, PPM(pixels per meter), observation requires 62 pixels, identification requires 125 pixels, and verification requires 250 pixels.
This standard is very popular in foreign countries, the good news is that the current domestic security manufacturers, such as Haikang Dahua, also introduced this standard in the introduction of product parameters.
Such as Dahua’s A certain camera callout DORI Distance:
detect:98.7 m(350.07 ft) Observe:39.5 m(129.59 ft) Recognize:19.7 m(64.63 ft) Identify:9.9 m(32.48 ft)
of course, there are similar standards in China. For example, GB/T 35678-2017 “Technical Requirements for Public Safety Face Recognition Application Images” stipulates that the distance between two eyes should be greater than or equal to 30 pixels, and should be greater than or equal to 60 pixels. The Ministry of Public Security GA/T893 “Security Biometrics Application Terms” stipulates that (for license plate recognition) letter recognition is 15 pixels (200 pixels/meter). The standards related to the face recognition algorithm stipulate that the face recognition requirement is 80 × 80 pixels.