Adaptive Noise Reduction According to Exposure, Gain and Image Level for Mobile Phone Cameras

This work proposes a de-noising scheme to restore images degraded by noise components resulting from the image sensors and image signal processors (ISP) used in mobile phone cameras. The noise model, measured using incident light values, is characterized by the exposure time , amplifier gain and image level. Thereafter, a three-step adaptive noise reduction and parallel concatenation filter techniques are implemented according to pre-defined noise model. In tests, a clearer image is restored and the noise signal reduced for all light values. The proposed filter techniques are suitable for implementation in low-power mobile phone cameras.