作者: Nitish V. Thakor , Wang Wei Lee , Haoyong Yu
DOI: 10.1109/BIOROB.2014.6913895
关键词: Exoskeleton 、 Sequence learning 、 Temporal resolution 、 Computer vision 、 Neuromorphic engineering 、 Computer science 、 Artificial intelligence 、 Pressure sensor 、 Standard deviation 、 Computational complexity theory 、 Bandwidth (signal processing)
摘要: We present a novel sampling and processing method for detecting gait events from an insole pressure sensor. Inspired by how tactile data is processed in the brain, we propose use of timing, instead intensity, as our event detection feature. By sacrificing need accurate intensity measurements, it possible to achieve superior temporal resolution, which arguably more important given timely feedback. In this paper, demonstrate temporally gait-event 1.2±7ms (mean standard deviation) heel-strike 0.2± 14ms toe-off compared reference system, success rate above 97% most trials, using only 1 bit information per channel. Our thus has potential much lower computational complexity bandwidth, both are key low-cost, portable solutions prosthetics, exoskeletons or long-term monitoring applications.