Recognition of user-dependent and independent static hand gestures: Application to sign language*

Static hand gesture (HG) recognition for both user-dependent and user-independent is a challenging problem, especially when there are changes in lighting, hand position, and background, the recognition becomes more complex. To solve this problem, this paper proposes a static hand gesture recognition based on a set of image descriptors: Gradient Local Auto-Correlation (GLAC), Gabor Wavelet Transform (GWT), and Fast Discrete Curve Transform (FDCT). Principal Component Analysis (PCA) was used to reduce dimensionality. Tests were performed on three sign language datasets and one hand posture dataset using neural network classifiers, K-Nearest Neighbor (KNN) classifiers, and combined classifiers. The results obtained were compared to the state of the art and show an accuracy of 100% for user-independent and 98.33% for user-dependent gestures, despite the difficult acquisition conditions of the datasets.