Hand Contour Detection Using Spatial Homomorphic Filtering and Variational Level Set

Hand contour detection is crucial for hand gesture recognition. Environmental illumination has a great influence on hand contour detection. In view of hand gesture images with nonuniform illumination, a practical and effective hand contour detection method is proposed. Firstly, in order to avoid frequency domain operations, a spatial homomorphic filtering is used to correct nonuniform illumination. Secondly, a variational level set method is used to detect hand contour. Experimental results demonstrate that the proposed method can detect hand contour accurately in the background with nonuniform illumination. The proposed method can be used in the fields of human computer interaction and augmented reality.