Segmentation-based adaptive spatio-temporal filtering for noise canceling and MPEG pre-processing

Noise reduction especially in low light level images is an important feature in consumer cameras. Existing methods to reduce such noise often degrade image quality due to an improper choice of filters. We present a high quality, low-cost noise reduction filter for enhancing low light level images. The proposed algorithm is simple, effective and computationally fast; it is suitable for low cost camcorders, digital cameras, CCTVs, and surveillance video systems.