Morse-code auto-recognition with Neural Network system for person with disabilities

This study uses Back Propagation Neural Network algorithm to improve the recognition rate of Morse code automatic recognition system. From a cerebral paralysis, the neural network has the average recognition rate up to 92%. The recognition rate from an amputee who used prosthesis for typing is 96.5% in average. The result is quite promising if comparing to 98.5% of the recognition rate from a skilled expert. Back Propagation neural network has successfully overcome the difficulty of analyzing severely unstable Morse code time series. It also makes possible to use neural network for real-time signal recognition since the human typing speed is quite slow in comparison to signal processing by the microprocessor.