The Improvement And Realization Of Automatic White Balance Algorithm

Digital image processing technology has been widely used in our daily life. AWB (Automatic White Balance) is a main method in correcting color-cast image, and AWB algorithms correct color-cast images by analysis color-cast image based on certain algorithm conjecture.AWB is widely used in digital devices, color scanners and digital processing software. Color cast images will be made out by the AO-format color scanner platform, when it runs in an unnormal illumination, or its CCD's not in good condition. The scanned image of Old books, pictures and maps which may be color cast, a can also be correct by AWB algorithms.In this paper, the AWB algorithms discussed in this paper are expected used in the AO-format color scanner platform. An improved white balance algorithm based on face color detection is presented, which has two important improvements. In face color detection section, a new face color detection method based on histogram matching is proposed which calculate a more reliably face color operator by getting rid of the color of hair, eyes, lips in face area. In correcting section, a new correcting method is discussed which improves correction by matching the GW white balance predetermined condition. It is proved that the new proposed algorithm has a wider usage and much better image-correcting results. Finally, an image correcting system is design to verify this algorithm, which is design in Matlab and Visual C++6.0 environment.