In this paper, Cellular Neural Networks using genetic algorithm (GA-CNNs) are designed for CMOS image noise reduction. Cellular Neural Networks (CNNs) could be an efficient way to apply to the image processing technique, since CNNs have high-speed parallel signal processing characteristics. Adaptive CNNs structure is designed for the reduction of Photon Shot Noise (PSN) changed according to the average number of photons, and the design of templates for adaptive CNNs is based on the genetic algorithm using real numbers. These templates are optimized to suppress PSN in corrupted images. The simulation results show that the adaptive GA-CNNs more efficiently reduce PSN than do the other noise reduction methods and can be used as a high-quality and low-cost noise reduction filter for PSN. The proposed method is designed for real-time implementation. Therefore, it can be used as a noise reduction filter for many commercial applications. The simulation results also show the feasibility to design the CNNs template for a variety of problems based on the statistical image model.

Image sensor noise was estimated in an approximately perceptually uniform space with a color image sensor model. Particularly, the noise level with respect to an image sensor's pixel pitch and the dark noise was investigated. It was shown that the noise level could be about half when spectral sensitivity was optimized considering noise with reduced color reproduction accuracy. It was also shown that for a 2.0 µm pixel pitch sensor, the exposure index should be less than 100-150 in order to keep the noise level 94 less than 5 even if it had no dark noise, whereas the exposure index could reach about 2000-4000 for a 8.0 µm pixel pitch sensor depending on the sensor sensitivity and the dark noise level.