Comparing camera performance in low light conditions
In this article,we will consider applications such as License Plate Recognition(LPR)or Optical Character Recognition(OCR),where black and white imaging is typically used,and the amount of light that a camera can collect may be limited by shorter exposure times.Determining the resolution,frame rate,and field of view required to solve imaging problems is relatively simple,but it is difficult to determine whether the camera has sufficient imaging capability.
This challenge is usually solved through repeated experimentation.Let's take an example:a visual system designer believes that a 1/4"CCD VGA camera running at a frame rate of 30FPS is sufficient for the aforementioned applications.Initial testing shows that the camera has sufficient sensitivity at an exposure time of 10ms when the object is stationary.The figure shows a simple example:characters B,8,D,and 0 are easily confused by visual algorithms.The image captured by a 1/4"CCD camera in the upper left corner is suitable for image processing.
However,when the object begins to move,the exposure time needs to be reduced,and the camera cannot provide useful information because the letters"B"and"D"cannot be distinguished from the numbers"8"and"0".In Figure 2,the images in the middle and bottom left show a degradation in image quality,especially those captured with a 1/4"CCD camera at an exposure time of 2.5ms,which is clearly not suitable for image processing.
In this example,assuming that a large depth of field is not required,the minimum F-value of the lens is acceptable.In other words,it is impossible to collect more light by opening the shutter of the lens.
Therefore,designers need to consider different cameras.The current question is whether choosing different cameras can improve the performance of the system.Using larger sensors has been widely recognized as a good solution to low light performance issues,so a 1/2"sensor would be a good choice.There is no need to explore further here,as the EMVA 1288 imaging performance of the reference camera is very useful.
Through EMVA 1288 data,it can be observed that a 1/4"CCD sensor has better quantum efficiency and lower noise;however,a 1/2"CCD sensor has larger pixels and saturation capacity.This article introduces how to determine whether a 1/2"camera has better performance.
The graph in Figure 3 compares the signal values and optical density(photon count/um2)of 1/4"and 1/2"cameras.The signal,as a function of optical density,can be determined by the following formula:
Signal value=optical density x(pixel size)2x quantum efficiency
An important assumption made in this article is that two cameras have the same settings,and the lenses have the same field of view and F value.
Figure 1 shows that under the same optical density,the 1/2"sensor will produce a higher signal.In addition,it can be observed in Figure 3 that both the 1/4"and 1/2"sensors have reached their saturation capacity at a light density of 700 photons/um2,but it is evident that the saturation capacity value of the 1/2"sensor is higher.
In the applications considered in this white paper,the comparison of cameras needs to be done at low light levels.Therefore,considering noise levels has become particularly important.
Figure 2 shows that the 1/2"sensor reaches an absolute sensitivity threshold at slightly lower optical densities compared to the 1/4"sensor.To further determine which camera performs better in low light applications,a more important measurement is the signal-to-noise ratio(SNR).
Figure 3 shows the functional relationship between SNR and optical density of two cameras
