Gait recognition based on the Fast Fourier Transform and SVM

Wang, Fei; Wen, Shengjun; Wu, Chengdong; Zhang, Yuzhong; Wang, Hao

doi:10.1109/ccdc.2011.5968347

Cited by 7 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each feature a is multiplied with a weight W. The multiplication of a and W is then added together with a bias b as stated in (17). The perceptron output y was scaled to a range of ±1 using a hyperbolic tangent sigmoid transfer function in (18)…”

Section: Standalone Gait Diagnosis Devicementioning

confidence: 99%

“…The proposed method is based on extracting some amplitudes of the 3D angles as features. In [17,18], Fourier transform of the gait data waveforms were used for gait classification. This is reasonable because Fourier transform could reconstruct all amplitudes of the 3D angles.…”

Section: Comparison Using Other Featuresmentioning

confidence: 99%

“…Mostayed et al [17] used the harmonic components of ankle-knee, knee-hip and hip-ankle angle waveforms to diagnose the gaits. Wang et al [18] used the Fourier transform of the shank's angular as a feature of SVM to recognise different activities such as walking and climbing upstairs.…”

Section: Introductionmentioning

confidence: 99%

“…Wang et al . [18] used the Fourier transform of the shank's angular as a feature of SVM to recognise different activities such as walking and climbing upstairs.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Comparison of machine learning methods for the construction of a standalone gait diagnosis device

Han¹,

Wong²,

Murray

2020

IET signal process.

View full text Add to dashboard Cite

In this research, the authors investigate the feasibility of selecting three-dimensional thigh and shank angles as the features of machine learning methods. Four common machine learning techniques, i.e. random forest, k-nearest neighbour, support vector machine and perceptron, were compared in terms of accuracy and memory usage so that a real-time standalone gait diagnosis device can be constructed using low-end inertial measurement units (IMUs). With proper re-sampling and normalisation, they discovered that the support vector machine and perceptron resulted in the top two highest accuracies (96-99%) among the four machine learning methods. The memory requirement of the perceptron is the lowest among the machine learning methods. Therefore, perceptron was selected as the classification algorithm for the standalone gait diagnosis device. The trained perceptron was transferred to the thigh and shank's IMUs to process the data locally in real-time. The constructed standalone gait diagnosis device lit up green or red light emitting diodes when normal or abnormal gaits were detected, respectively. This standalone device was further tested in real-life and achieved a mean classification accuracy of 96.50%. AccGy stands for acceleration and gyroscope. The extracted features are similar to that of Mikos et al. [4]. DFT stands for discrete Fourier transform. The features are the first 5 harmonic components (5 instantaneous amplitudes and 5 phases) of the angle waveform. These results were obtained using all features extracted from 3D thigh and shank sensor data. Values in bold font means ≥ 96%.

show abstract

Section: Standalone Gait Diagnosis Devicementioning

confidence: 99%

Section: Comparison Using Other Featuresmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…Wang et al . [18] used the Fourier transform of the shank's angular as a feature of SVM to recognise different activities such as walking and climbing upstairs.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Comparison of machine learning methods for the construction of a standalone gait diagnosis device

Han¹,

Wong²,

Murray

2020

IET signal process.

View full text Add to dashboard Cite

show abstract

“…Because of the important information contained in gait cycle, researchers have proposed the method to extract the gait signal from the accelerometer sensor. Fei Wang utilized fast Fourier transform (FFT) and support vector machine (SVM) for feature selection and gait recognition [6]. Dong Xu implemented matrix-based algorithm: couple subspace analysis (CSA) and discriminant analysis for human gait recognition [7].…”

Section: Introductionmentioning

confidence: 99%

Step Estimator Based on a Wearable ECG Monitor

Prawiro

Lin

2018

Mobile Information Systems

View full text Add to dashboard Cite

ECG signal acquisition may be contaminated with various unwanted signals from the environment, but the contaminating signals are not always be categorized as noise. In this paper, we use the signal processing method to extract the step signal from an ECG signal. The proposed algorithm has been designed, tested, and evaluated using ensemble empirical mode decomposition (EEMD) algorithm as ECG data extraction and accelerometer data as the step counter reference. The main idea is to choose the correct intrinsic mode function (IMF) that represents gait signal delineation of ECG, and then calculating the detection accuracy of gait signal based on the step counter from accelerometer data. We use different belt tightness configurations based on chest circumference for each subject during the treadmill test, and then perform the detection accuracy for each configuration. Five healthy subjects were participated in the treadmill test as training subjects. Each participant was asked to walk in four levels of speeds and continue to run in five different speeds: 1.8, 2.7, 3.6, 4.5, 5.4, 6.3, 7.2, 8.1, and 9.0 km/h. Later on, other five healthy subjects were used to evaluate the accuracy of the proposed algorithm using identical configuration as training subjects. The proposed algorithm is able to achieve the detection accuracy of 91.93% and 94.72% for training and testing subjects using the belt tightness scenario of 100% length of chest circumference.

show abstract