Md. Rezwanul Ahsan

Hand motion detection from EMG signals by using ANN based classifier for human computer interaction

Todays advanced muscular sensing and processing technologies have made the acquisition of electromyography (EMG) signal which is valuable. EMG signal is the measurement of electrical potentials generated by muscle cells which is an indicator of muscle activity. Other than rehabilitation engineering and clinical applications, EMG signals can also be employed in the field of human computer interaction (HCI) system. In this work, the detection of different hand movements (left, right, up and down) was obtained using artificial neural network (ANN). A back-propagation (BP) network with Levenberg-Marquardt training algorithm was utilized. The conventional time and time-frequency based feature sets have been chosen to train the neural network. The simulation results show that the designed network is able to recognize hand movements with satisfied classification efficiency in average of 88.4%.

Neural network classifier for hand motion detection from EMG signal

EMG signal based research is ongoing for the development of simple, robust, user friendly, efficient interfacing devices/systems for the disabled. The advancement can be observed in the area of robotic devices, prosthesis limb, exoskeleton, wearable computer, I/O for virtual reality games and physical exercise equipments. Additionally, electromyography (EMG) signals can also be applied in the field of human computer interaction (HCI) system. This paper represents the detection of different predefined hand motions (left, right, up and down) using artificial neural network (ANN). A back-propagation (BP) network with Levenberg-Marquardt training algorithm has been utilized for the classification of EMG signals. The conventional and most effective time and time-frequency based feature set is utilized for the training of neural network. The obtained results show that the designed network is able to recognize hand movements with satisfied classification efficiency in average of 88.4%. Furthermore, when the trained network tested on unknown data set, it successfully identify the movement types.

Bandwidth Enhancement of a Dual Band Planar Monopole Antenna Using Meandered Microstrip Feeding

A meandered-microstrip fed circular shaped monopole antenna loaded with vertical slots on a high dielectric material substrate (ε r = 15) is proposed in this paper. The performance criteria of the proposed antenna have been experimentally verified by fabricating a printed prototype. The experimental results show that the proposed antenna has achieved wider bandwidth with satisfactory gain by introducing meandered-microstrip feeding in assistant of partial ground plane. It is observed that, the −10 dB impedance bandwidth of the proposed antenna at lower band is 44.4% (600 MHz–1 GHz) and at upper band is 28% (2.25 GHz–2.95 GHz). The measured maximum gains of −1.18 dBi and 4.87 dBi with maximum radiation efficiencies have been observed at lower band and upper band, respectively. The antenna configuration and parametric study have been carried out with the help of commercially available computer-aided EM simulator, and a good accordance is perceived in between the simulated and measured results. The analysis of performance criteria and almost consistent radiation pattern make the proposed antenna a suitable candidate for UHF RFID, WiMAX, and WLAN applications.

Low-Cost Dielectric Substrate for Designing Low Profile Multiband Monopole Microstrip Antenna

This paper proposes a small sized, low-cost multiband monopole antenna which can cover the WiMAX bands and C-band. The proposed antenna of 20 × 20 mm2 radiating patch is printed on cost effective 1.6 mm thick fiberglass polymer resin dielectric material substrate and fed by 4 mm long microstrip line. The finite element method based, full wave electromagnetic simulator HFSS is efficiently utilized for designing and analyzing the proposed antenna and the antenna parameters are measured in a standard far-field anechoic chamber. The experimental results show that the prototype of the antenna has achieved operating bandwidths (voltage stand wave ratio (VSWR) less than 2) 360 MHz (2.53–2.89 GHz) and 440 MHz (3.47–3.91 GHz) for WiMAX and 1550 MHz (6.28–7.83 GHz) for C-band. The simulated and measured results for VSWR, radiation patterns, and gain are well matched. Nearly omnidirectional radiation patterns are achieved and the peak gains are of 3.62 dBi, 3.67 dBi, and 5.7 dBi at 2.66 GHz, 3.65 GHz, and 6.58 GHz, respectively.

The Use of Artificial Neural Network in the Classification of EMG Signals

This paper presents the design, optimization and performance evaluation of artificial neural network for the efficient classification of Electromyography (EMG) signals. The EMG signals are collected for different types of volunteer hand motion which are processed to extract some predefined features as inputs to the neural network. The time and time-frequency based extracted feature sets are used to train the neural network. A back-propagation neural network with Levenberg-Marquardt training algorithm has been employed for the classification of EMG signals. The results show that the designed and optimized network able to classify single channel EMG signals with an average success rate of 88.4%.

VHDL Modelling of Fixed-point DWT for the Purpose of EMG Signal Denoising

Wavelet Transform (WT) has been widely applied in biomedical signal analysis. This paper will present the denoising method of EMG signal using WT and its model processed by VHSIC (Very High Speed Integrated Circuit) Hardware Description Language (VHDL) model of it. The principle of wavelet denoising is first to decompose the signal by performing a WT, followed by applying suitable thresholds to the detail coefficients, zeroing all coefficients below their associated thresholds, and finally to reconstruct the denoised signal based on the modified detail coefficients. Discrete Wavelet Transform (DWT) is a method that uses wavelet analyzer in which case the signal split into small pieces preserving both time and frequency properties. The Second order of Daubechies family (db2) has been used to denoise EMG signals. The simulation, synthesis and verification of the design presents a fast and reliable prototyping of DWT for denoising of EMG signals.

Analysis of a Compact Wideband Slotted Antenna for Ku Band Applications

The design procedure and physical module of a compact wideband patch antenna for Ku band application are presented in this paper. Finite element method based on 3D electromagnetic field solver has been utilized for the designing and analyzing process of proposed microstrip line fed modified - shaped electrically small patch antenna. After successful completion of the design process through various simulations, the proposed antenna has been fabricated on printed circuit board (PCB) and its characteristics have been studied. The parameters of the proposed antenna prototype have been measured in standard far-field rectangular shape anechoic measurement compartment. It is apparent from the measured antenna parameters that the proposed antenna achieved almost stable variation of radiation pattern over the entire operational band with 1380 MHz of 10 dB return loss bandwidth. The maximum gain of 7.8 dBi and 89.97% average efficiency within the operating band from 17.15 GHz to 18.53 GHz ensure the suitability of the proposed antenna for Ku band applications.

A Dual Band Slotted Patch Antenna on Dielectric Material Substrate

A low profile, compact dual band slotted patch antenna has been designed using finite element method-based high frequency full-wave electromagnetic simulator. The proposed antenna fabricated using LPKF printed circuit board (PCB) fabrication machine on fiberglass reinforced epoxy polymer resin material substrate and the performance of the prototype has been measured in a standard far-field anechoic measurement chamber. The measured impedance bandwidths of (reflection coefficient  dB) 12.26% (14.3–16.2 GHZ), 8.24% (17.4–18.9 GHz), and 3.08% (19.2–19.8) have been achieved through the proposed antenna prototype. 5.9 dBi, 3.37 dBi, and 3.32 dBi peak gains have been measured and simulated radiation efficiencies of 80.3%, 81.9%, and 82.5% have been achieved at three resonant frequencies of 15.15 GHz, 18.2 GHz, and 19.5 GHz, respectively. Minimum gain variation, symmetric, and almost steady measured radiation pattern shows that the proposed antenna is suitable for Ku and K band satellite applications.

Compact double-p slotted inset-fed microstrip patch antenna on high dielectric substrate.

This paper presents a compact sized inset-fed rectangular microstrip patch antenna embedded with double-P slots. The proposed antenna has been designed and fabricated on ceramic-PTFE composite material substrate of high dielectric constant value. The measurement results from the fabricated prototype of the antenna show −10 dB reflection coefficient bandwidths of 200 MHz and 300 MHz with center resonant frequency of 1.5 GHz and 4 GHz, respectively. The fabricated antenna has attained gains of 3.52 dBi with 81% radiation efficiency and 5.72 dBi with 87% radiation efficiency for lower band and upper band, respectively. The measured E- and H-plane radiation patterns are also presented for better understanding. Good agreement between the simulation and measurement results and consistent radiation patterns make the proposed antenna suitable for GPS and C-band applications.

Optimization of neural network for efficient EMG signal classification

This paper illustrates the classification of Electromyography (EMG) signals through designing and optimization of artificial neural network. The EMG signals obtained for different kinds of hand movements, which are processed to extract the features. Extracted time and time frequency based feature sets are used to train the neural network. A back-propagation neural network with Levenberg-Marquardt training algorithm has been utilized for the classification. The results show that the designed network is optimized for 10 hidden neurons and able to efficiently classify single channel EMG signals with an average rate of 88.4%.