Low Contrast Enhancement for Facial Feature Expression

Detection of facial features in low lighting conditions is a major challenge in the development of computer visionbased monitoring systems, such as driver drowsiness detection. To address this problem, this research proposes an illumination invariant image enhancement method to improve the quality of facial picture in low light conditions. This method incorporates the Retinex model based on adaptive gamma correction and color saturation mapping, and the Interval Valued Intuitionistic Fuzzy Sets (IVIFS) based fuzzy to enhance contrast and preserve important details in the facial area. The enhancement process is performed through RGB to HSV color space conversion, luminance component enhancement, and color saturation adjustment. The enhanced image is then used to extract visual features such as eye-mouth aspect ratio and head pose, which are crucial in drowsiness detection systems. Experimental results show that this method is able to improve the quality of facial feature detection under low lighting conditions.