A Compressed Sensing Based Method for Reducing the Sampling Time of A High Resolution Pressure Sensor Array System

Abstract

For extracting the pressure distribution image and respiratory waveform unobtrusively and comfortably, we proposed a smart mat which utilized a flexible pressure sensor array, printed electrodes and novel soft seven-layer structure to monitor those physiological information. However, in order to obtain high-resolution pressure distribution and more accurate respiratory waveform, it needs more time to acquire the pressure signal of all the pressure sensors embedded in the smart mat. In order to reduce the sampling time while keeping the same resolution and accuracy, a novel method based on compressed sensing (CS) theory was proposed. By utilizing the CS based method, 40% of the sampling time can be decreased by means of acquiring nearly one-third of original sampling points. Then several experiments were carried out to validate the performance of the CS based method. While less than one-third of original sampling points were measured, the correlation degree coefficient between reconstructed respiratory waveform and original waveform can achieve 0.9078, and the accuracy of the respiratory rate (RR) extracted from the reconstructed respiratory waveform can reach 95.54%. The experimental results demonstrated that the novel method can fit the high resolution smart mat system and be a viable option for reducing the sampling time of the pressure sensor array.

Compressive sensing based maximum-minimum subband energy detection for cognitive radios

“…When the number of measurements for spectrum sensing is reduced, the time required for SU in the sampling process is also reduced significantly. Therefore, the time consumed by SU during spectrum sensing increases along with the increasing number of measurements [45]. Additionally, a shorter sensing time during SS has a longer data transmission duration that increases the throughput of the SUs.…”

“…The time of acquiring and sending both the spectrum values and their locations information for 5000 points is near the time of acquiring every one of the 10000 spectrum values using the traditional sampling method governed by the Shannon-Nyquist sampling theorem [45]. Therefore, beginning from 5000 measurements, every reduction of 2000 measurement will lead to the 40% ( 10000−6000 10000 * 100%…”

An Unobtrusive and Non-Contact Method for Respiratory Measurement With Respiratory Region Detecting Algorithm Based on Depth Images

“…The three different p values in the second step were selected based on empirical experiments, and it can effectively solve the problem that the predefined smoothing parameter was incapable to be applied in various respiratory patterns with different RRs. Then define the average accuracy of RR for estimating the performance of generating the depth waveform in the following form [21]:…”

Continuous and rapid fabrication of photochromic fibers by facilely coating tungsten oxide/polyvinyl alcohol composites

“…Modern clothing is currently moving towards more and more intelligent applications with the addition of some high-end functionalities, such as physiological signal acquisition, 1 performance monitoring, 2 flexible electronics and display, 3 temperature control, 4 energy storage, 5 etc. However, most of the current so-called “smart clothing” is made of technologies being integrated into common garments rather than made of truly functional fibers.…”