A Novel Hand Gesture Recognition Method Based on Illumination Compensation and Grayscale Adjustment

Gesture recognition is a challenging research problem in human–machine systems. Uneven illumination and background noise significantly contribute to this challenge by affecting the accuracy of hand gesture recognition algorithms. To address this challenge, this paper proposes a novel gesture recognition method based on illumination compensation and grayscale adjustment, which can significantly improve gesture recognition in uneven and backlighting conditions. The novelty of the method is in the new illumination compensation algorithm based on luminance adjustment and Gamma correction, which can reduce the luminance value in the overlit image region and enhance the area with low illumination intensity. The grayscale adjustment is used to detect the skin color and hand area accurately. The binary image of the hand gesture is extracted through iterative threshold segmentation, image dilation, and erosion process. Five gesture features including area, roundness, finger peak number, hole number, and average angle are used to recognize the input gesture. The experimental results show that the proposed method can reduce the influence of uneven illumination and effectively recognize the hand gestures. This method can be used in applications involving human–machine interactions conducted in poor lighting conditions.