Rotation-invariant Face Detection in Color Images with Complex Background
Jingyu Yang · Jisuanji gongcheng · 2008
This paper presents a method for face detection in color images with complex backgrounds. Which adopts color balance method to correct the images color bias in the RGB color space based on Gray World hypothesis; detects skin color in the YES color space, introduced preprocessing techniques to reduce the searching areas of the face detection, and proposes a method that can detect arbitrary rotated faces based on the computation of the direction of the object. Template-matching method is used in the face candidate regions to locate face. The experiments show that the proposed face detection method provides better detection results to different illumination, different sizes and arbitrary rotated faces.