Image Pipeline Tuning for Digital Cameras

Abstract

Digital cameras are widely used in many applications, such as digital still cameras, camcorders, camera phones, and video surveillance. The image pipeline plays a key role in digital camera systems by generating a digital color image in real time from raw data captured by a camera sensor. However, tuning an image pipeline for a particular camera system is a challenging task because different lenses, camera modules, and sensors have different characteristics. Moreover, there are no standard image quality metrics and different customers have different preferences for image quality. We present basic image pipeline algorithms and tuning methods from a system point of view.