HLS based hardware acceleration on the zynq SoC: A case study for fall detection system

Abstract

Fall detection is a major problem in healthcare systems, especially for elderly people who are the most vulnerable. It is important to design and implement not only an accurate fall detection system (FDS) but also a system with a real-time response. The achievement of high accuracy and fast response time together allows the development of a system that helps saving lives, time and money in healthcare industry. This paper presents the design, simulation and implementation of a novel FDS using the Shimmer wearable sensor. The discrete wavelet transform (DWT) is applied for preprocessing the data coming from the Shimmer platform, principal component analysis (PCA) is used for dimensionality reduction and feature extraction and finally, a binary decision tree (DT) is utilized for classification purpose. The system is simulated in MATLAB prior to the implementation on the Zynq system-on-chip (SoC) for hardware acceleration. DWT is executed on the processing system (PS) of the Zynq platform in a software manner while PCA and DT are both implemented on the programmable logic (PL) for hardware acceleration. PCA and DT are developed in C and synthesized in Vivado high level synthesis (HLS) tool to transform the C based designed into a register transfer level (RTL) implementation. Various optimization techniques are explored in Vivado HLS. The performance of the FDS in terms of accuracy of the classifier is 88.4% while the overall resources used in PL of the Zynq vary between 2% and 23% depending on the running frequency and optimization technique used.

Triaxial Accelerometer-Based Falls and Activities of Daily Life Detection Using Machine Learning

“…The recorded triaxial accelerometer signal consisted of a 3-dimensional vector, with each component corresponding to each of the three axes/dimensions x , y , z respectively. An analysis of the captured signals showed that the duration of an event that caused acceleration was not longer than 2 s, as also pointed out in [ 26 ]. Using the information from the captured videos from each experiment and the corresponding video and accelerometer timestamps, the acquired signals were first divided into segments that contained only one event each.…”

“…Ali et al [ 26 ] proposed a fall detection system based on a system-on-chip (SoC) board that utilised a triaxial accelerometer for real-time fall detection. The Discrete Wavelet Transform (DWT) and PCA were used for feature extraction and fall detection accuracy reached 88.4% using a Decision Tree classifier.…”

Hardware/Software Co-Design of Fractal Features Based Fall Detection System

“…The processing of entire window segment with arithmetic trees results in a higher throughput and an effective latency of 0.1 µs/sample with 128 sample bursts of data streaming into the proposed hardware design. The latest works in [66,67] have a higher latency of 1 s, while [47,50,51] have latencies in microseconds, however they suffer form lower classification performance. Furthermore, the proposed system also consumes relatively low to moderate resources over all hardware implementations from design I to final design, which allows flexibility in performance and area trade-off requirements for processing.…”

“…However, their implementation did not give any power consumption values. Ali et al [47] implemented a sensor-based FDS on a Zynq System on Chip (SoC) device which applies DWT and PCA for feature extraction and a binary decision tree for classification. The system suffered from lower accuracy and no power consumption analysis has been performed.…”

Fall Detection FPGA-Based Systems: A Survey

“…The authors in [48] present the design, simulation and implementation of a novel fall detection system based on the Shimmer platform and Zynq board. In this prototype, accelerometer data are sent to the processing unit by Bluetooth.…”