Real Time Palm and Finger Detection for Gesture Recognition Using Convolution Neural Network

Hand motion is one of the methods for communicating with the PC to perform colossal application. The goal is to apply the machine learning algorithm for quicker motion route acknowledgment. This application pursues the learning of motions and recognizing them accurately. At first Convolution Neural Network (CNN) model is prepared with various pictures of various hand signals for different people for grouping the letters in order. The initial step is realizing, where the data is stacked and in pre-handling step, edge, mass locators, twofold thresholding, limit box recognitions are used to extract the features. By utilizing ConvNet, which is a machine learning algorithm, the input picture’s features are cross confirmed. The accuracy is found to be 98% and this methodology is effective to address the impact of various problems. The assessment of this strategy shows to performing hand gesture acknowledgment.