An Improved Hand Gesture Recognition System Based on Optimized MSVM and SIFT Feature Extraction Algorithm

In today, the robots and machine, translation tasks are performing the main role in hand gestures. Gesture or detection implementation will help humans in several means. Gesture recognition systems are used in various fields such as DNN, ML, and NNs. The applications of this research are sign language translation, music creation, robot remote controlling, etc. In this research work, proposed HGR methods are using feature extraction and optimized MSVM classification. This process gives the applications of edge detection, interference, filters, SIFT-ALO algorithm, and binary during image preprocessing, in which these approaches add to better extraction and selection. In proposed work, implemented an optimized MSVM method are two classes such as training and testing. Optimized M-SVM is performed on the ASL gesture dataset along with existing SURF and SIFT techniques. This research work is used for the simulation tool (MATLAB) and calculated performance metrics like processing time, error rate, and the accuracy rate with the MSVM value at 99.1 percent as compared to existing feature extraction methods.