Utilizing gammatone filter coefficient to improve human mouth-click signal detection using a multi-phase correlation process

Abstract

This study introduces an enhanced structured signal processing scheme for detecting the “transmit-echo” of mouth-click signals used by blind individuals for echolocation. The processing scheme is based on coexisting work on the detection of the mouth-click signal, with an additional band-pass filter process introduced before synthesis, multi-phase correlation, and summation. The level of side lobes at the output was improved by more than −19 dB, and the number of local maxima was minimized by using a band-pass filter. The detection of “transmit-echo” results using artificial mouth-click signal data was validated and achieved a 100% success rate in detecting obstacles at 60 cm, 80 cm, and 100 cm. Nonetheless, the detection scheme discussed in this investigation is thought to be intuitive, having been learned from the human hearing process. The emerging concepts in this research are expected to benefit radar and sonar system applications in the near future.