Camera Image Processing Principle Analysis Anti-noise Zoom Frequency Brightness Sensing and Exposure Camera Image Processing Principle Analysis - Anti-noise Zoom Frequency Brightness Sensing and Exposure 1.1 Anti-noise Processing The increase of AG will inevitably lead to an increase in noise. In addition, if the light is dim and the exposure time is too long, the number of noise will also increase (from the perspective of digital cameras, it is mainly because of long exposure, the temperature of the photosensitive element increases, and the current noise causes the increase of noise in the photosensitive element). The defects of the photosensitive element itself are also one of the sources of noise or even bad pixels. Therefore, sensor integration or back-end ISP usually have related settings for noise reduction. 1.1.1 Startup Timing According to the cause of noise formation, the noise reduction function needs to be started after AG or Exptime exceeds a certain value. Therefore, it is usually necessary to determine the threshold of these two parameters. Too small or too large is not good. From the following noise reduction processing method, you will see that noise reduction will bring about a decrease in image quality. Therefore, starting the noise reduction function too early and doing noise reduction processing when it is unnecessary will not only increase the burden on the processor or ISP, but may also be counterproductive. If the noise reduction function is activated too late, it will not work when it is needed. 1.1.2 Judgment principle and processing method How to judge whether a point is a noise point? Let\'s start with how people identify noise points. For the human eye, judging whether a point is a noise point is nothing more than that the brightness or color of this point is too different from most of the points on the edge. From the mechanism of noise generation, color abnormality should always be accompanied by brightness abnormality, and the workload of processing brightness abnormality is smaller than that of color abnormality. Therefore, the judgment principle of sensorISP is usually that when the difference between the brightness of a point and the brightness of the surrounding points is greater than a threshold, the point is considered to be a noise point. The processing method is usually to take the average of the surrounding points to replace the original value. This approach does not increase the amount of information and is similar to a fuzzy algorithm. For high-end digital cameras with powerful image processing chips, whether there are more complex algorithms in judgment and processing is also possible. For example, brightness and color are used as the standard to judge noise, and an interpolation algorithm with a larger amount of calculation is used to compensate. For the inherent bad pixels and noise of the sensor, the data is discarded by shielding (Nikon does this...
You Might Like
Recommended ContentMore
Open source project More
Popular Components
Searched by Users
Just Take a LookMore
Trending Downloads
Trending ArticlesMore