A Novel Fuzzy Logic Based Hand Gesture Recognition System

In this proposed study, a Fuzzy Logic System (FLS) was developed to classify and detect hand movements. The designed FLS system consists of a Fuzzifier, Inference Engine, Knowledge Base, and Defuzzifier. The Mamdani technique was used as the Inference Engine, and the centroid method was used for Defuzzification. Five input variables (Flex1-5) and one output variable (Sign) were used to create a rule base with 94 rules. A sensor array was placed on a glove to generate data, and a data collection circuit was established. Movements were performed through this circuit to create the rule bases. A total of 15.030 data points were analyzed to develop the FLS. According to the results, the movements (97.5%) were detected successfully.