An Evolutionary Approach to Digitalized Hand Signs Recognition

The issue of automatically recognizing digitalized human-made hand signs is a crucial step in facing human-computer interaction and is of paramount importance in fields such as domotics. In this paper Differential Evolution is used to perform classification of hand signs collected in a reduced version of the Auslan database. The performance of the resulting best individual is computed in terms of error rate on the testing set and is compared against those of other ten classification techniques well known in literature. Results show the effectiveness and the efficacy of the approach in solving the recognition task.