Face detection under varying lighting conditions in video streams

Auto human face detection from video streams is the base of studies for human face recognition and tracking. This paper proposes an efficient and robust method to detect face in video streams based on image enhancement, Gabor wavelet transform and adaboost algorithm. The key step and the main contribution of this work is to use the technique of image enhancement to alleviate the impact of human face detection caused by variation illumination such as local shadow and highlight. The approach uses a cascade of classifiers to adopt a coarse-to-fine strategy for achieving higher detection rates with lower false positives. The experimental results demonstrate that our proposed approach can improve the accuracy of face detection significantly even under varying lighting conditions.