Deep learning based static hand gesture recognition

Hand gesture recognition becomes a popular topic of deep learning and provides many application fields for bridging the human–computer barrier has positive impact on our daily life. The primary idea project is static acquisition from depth camera to process input images train convolutional neural network pre-trained ImageNet dataset. Proposed system consists capture device (Intel® RealSense™ D435), pre-processing image segmentation algorithms, feature extraction algorithm object classification. For algorithms computer vision methods OpenCV Intel Real Sense libraries are used. subsystem features extracting gestures classification based modified VGG-16 by using TensorFlow&Keras framework. Performance evaluated maching metrics. Experimental results show that proposed model, trained database 2000 images, high accuracy both at training testing stages.