The main parameters of industrial cameras include resolution, frame rate, signal-to-noise ratio, pixel depth, etc. In addition, other important parameters also include dynamic range.
Dynamic range refers to the brightness range in the scene or the photosensitive range of photosensitive materials such as film, CCD, CMOS, etc. For film and photosensitive components, dynamic range represents the range from "darkest" to "brightest" contained in the image. The larger the dynamic range, the richer the levels that can be represented, and the wider the color space it contains. So, what happens when the dynamic range of the shooting scene exceeds the dynamic range of the camera.
The photosensitive components of industrial cameras are composed of millions of pixels, which absorb photons during pixel exposure, convert them into digital signals, and then image. This process is like taking millions of buckets outdoors to collect rainwater. The brighter the photosensitive area, the more photons are naturally collected. After exposure, the photosensitive components are assigned discontinuous values based on the amount of photons collected for each pixel, and converted into digital signals. The pixel values that do not absorb photons and absorb photons to full load are displayed as "0" and "255" respectively (using an 8-bit image as an example), representing pure black and pure white.
Once these pixels are fully loaded, photons will overflow, causing loss of information (details). Taking red as an example, highlight overflow causes the values of other pixels near the pixel filled with red to become 255, but in reality, their true values have not reached 255. In other words, the details of the image are lost, which can result in missing information in the highlights. If we reduce the exposure time to prevent highlight overflow, many pixels used to describe dim environments do not have enough time to receive photons, resulting in a pixel value of 0, which can lead to missing information in the dim part.
The dynamic range of real-world scenes can reach 1014 orders of magnitude, and the human visual system (HVS) can perceive a dynamic range of 105 orders of magnitude. Through the adaptive mechanism of the human visual system, it can perceive a dynamic range of 109 orders of magnitude. However, the dynamic range of current digital image acquisition and display devices is relatively limited, with only 102-103 orders of magnitude. That's why current industrial cameras cannot directly capture images with HDR effects, and HDR technology can be further developed and applied.
So, how to improve the dynamic range of industrial cameras in the actual shooting process?
1. Using large-sized CCD targets. From the above explanation, we can now understand why digital cameras using large-sized photosensitive components have a larger dynamic range. The reason is simple: the size of the photosensitive components in digital SLRs is generally 4-10 times that of consumer grade digital cameras, allowing them to carry more pixels without reducing the distance between pixels and causing noise. More pixels will not be quickly "filled", so pixels representing dim environments have more time to absorb photons before pixels representing bright environments are fully loaded, resulting in richer image details. Therefore, in industrial testing, we tend to use larger format cameras.
2. Set the exposure settings for the sensor, automatically using long-term exposure in low light conditions and short-term exposure in high light conditions. This method can cause changes in the proportion of grayscale values of the corresponding objects in the image. If used for image processing, it cannot truly obtain effective information of the original scene. Therefore, this method cannot be used in industrial cameras, which is also the reason why most industrial cameras do not have the high dynamic range of civilian cameras.
