Title :
A linear filtering solution for color filter array interpolation
Author :
Glotzbach, John W. ; Schafer, Ronald W.
Abstract :
Digital cameras measure a color image by using color filters to interleave red, green, and blue samples across the charge coupled device (CCD) sensor. These samples must be interpolated to the resolution of the sensor to produce a full-color image. Adams and Hamilton proposed an interpolation method for the green image where for each missing sample, an estimate was computed by interpolating along a vertical or horizontal direction. In this method, red, green, and blue samples were used to estimate the green value. This paper provides an analysis of this method and simplifies the algorithm by removing the directional processing. The new method uses one interpolation filter for the green pixels and one interpolation filter for the red and blue pixels. Because the proposed method is a linear system, it runs efficiently on a DSP system. The quality of the output images from the proposed method is equivalent to the quality of the images from the Adams and Hamilton method.
Keywords :
filtering theory; image colour analysis; image resolution; interpolation; linear systems; blue image; charge coupled device sensor; color filter array interpolation; color filters; color image; directional processing; green image; interpolation method; linear filtering solution; linear system; red image; Charge coupled devices; Charge measurement; Charge-coupled image sensors; Color; Current measurement; Digital cameras; Digital filters; Interpolation; Maximum likelihood detection; Nonlinear filters;
Conference_Titel :
Digital Signal Processing Workshop, 2002 and the 2nd Signal Processing Education Workshop. Proceedings of 2002 IEEE 10th
Print_ISBN :
0-7803-8116-5
DOI :
10.1109/DSPWS.2002.1231088