IMAGE PIPELINE ALGORITHMS FOR STANDARD MOBILE IMAGING ARCHITECTURE SENSORS
Bo-Ching Huang, Chiou‐Shann Fuh · 2005
With rapid growth of DSC and camera phones, image pipeline becomes more and more important. Besides, new standard between camera module and mobile host called SMIA (Standard Mobile Imaging Architecture) are proposed last year. We will briefly introduce the standard in this paper. Image pipeline is a technique to transform sensor’s raw data to final compressed image. The images are processing through the pipeline: Optical Black Clamp, White Balance, Color Interpolation, Color Correction, Gamma Correction, Edge Enhancement, and Image Compression. This thesis contains original thought in some steps of the image pipeline. In White Balance, we propose a new method called Modified Gray World Assumption (MGWA). In Color Interpolation, we combine gradient and Laplacian edge detectors to do Edge-Sensitive Interpolation. In Edge Enhancement, we proposed a new method called Dynamic High-Boost Filtering. Also, we bring up an idea that Saturation Adjustment can be added in the image pipeline algorithm.