Research on gesture recognition preprocessing technology based on skin color detection

In the process of gesture recognition, human hand detection based on skin color is a method with high recognition rate.When the acquired gesture information is converted into digital images that can be processed by computer, the gesture images will be interfered and affected by various factors in the process of generation, transmission or transformation, and the quality of the gesture images will be distorted to different degrees due to noise, so the image needs to be preprocessed first.The process of image preprocessing includes image binarization and image smoothing. In this paper, the smoothing technology and binarization method are respectively studied and compared in experiments, providing effective data samples for gesture segmentation and feature extraction in gesture recognition.