International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 08 | Aug 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5003 Detection of Stress Level of Automobile Drivers using ECG and EMG Signals Miss. Dikshita D. Sheth 1 , Dr. Sanjay L. Nalbalwar 2 , Dr. Shankar B. Deosarkar 3 1 M. Tech Scholar, Dept. of Electronics and Telecommunication Engineering, Dr. Babasaheb Aambedkar Technological University, Lonere, Maharashtra- India 2 Associate Professor, Dept. of Electronics and Telecommunication Engineering, Dr. Babasaheb Aambedkar Technological University, Lonere, Maharashtra- India 3 Associate Professor, Dept. of Electronics and Telecommunication Engineering, Dr. Babasaheb Aambedkar Technological University, Lonere, Maharashtra- India ---------------------------------------------------------------------***---------------------------------------------------------------------- Abstract – In the past few decades, there was a steep increase in road accidents and also loss of life was experienced due to increase in the driver's mental stress. Thus stress level detection is very important in automobile drivers. In this paper, a method is proposed to detect stress level using features extracted from biomedical signals such as ECG and EMG signals. These signals were taken from “Physionet Database” and the results are obtained using MATLAB Simulation Software. The extracted features showed correlation with the stress level of automobile drivers, with the high accuracy of 78.57% for 60 seconds of the signals and 92.85% for 240 seconds of the signals. Key Words: Signal Processing, ECG, EMG, Stress Detection, Automobile Driver, MATLAB. 1. INTRODUCTION Driving could be a complex task which needs full concentration and an important balance between the alertness and a relaxed attitude [1]. Stress and powerful emotions can affect this balance whether or not they result from the driving task itself, or some unrelated matters. The work of operating transportation system vehicles in urban centers is also among the foremost stressful and unhealthy modern occupations. According to survey conducted by Cigna TTK insurance, about 89% of India’s population stricken by stress, as compared to the worldwide average of 86%. A survey conducted by Ford Motor Company about Causes of Stress in the Indian car drivers states that most of the drivers in India suffer from stress due to congested traffic. About 63% drivers feel stress due to traffic jams, while 56% of the Indian drivers suffer from the stress due to improper management of parking. 45% drivers fear of getting into an accident while 37% drivers fears due to increasing fuel prices. An ECG can be defined as a biological signal which represents the electrical activity of the heart associated which is with oscillation. It can be measured as a potential difference using special electrodes placed on different areas of the chest and limbs [2]. Heart rate can give an appropriate state of the human stress. Many studies have proven the connection between the HRV (heart rate variability) and the stress level [3]. Feature extracted from the ECG signal for the proposed method in this paper are R-peaks and Average Heart Rate. An EMG signal is the one of the most significant biomedical signals that demonstrates stress of a person. The EMG signal is random in nature. Feature extracted from the ECG signal for the proposed method in this paper is ZCR (Zero Crossing Rate). When a driver suffers from the low stress, the action potential propagates a chemical reaction that creates the relaxation of muscle fibres. Hence get low ZCR value. While when muscles are in contracted condition, ZCR value becomes higher. 2. LITERATURE REVIEW The proposed method states the stress level detection of different automobile drivers under different traffic conditions. Different approaches used by various researchers are tabulated as follow: Table -1: Various approaches by other researchers References Algorithm used Physiological signals used Populations used Picard & Healey (2000)[4] Linear Discrimination Algorithm ECG, EMG, SC (Skin Conductivity), Resp. 3 experienced drivers Karthik S, Sathiya A, Suganthi N (2014)[5] SVM (Support Vector Machines) ECG,EMG 14 individuals Barreto& Zhai (2006)[6] SVM (Support Vector Machines) BVP, GSR, PD 32 individuals Singh R. R. & Banerjee R.(2013)[7] LRNN EDA, HRV, RSP signals 19 individuals The various researchers previously used various features are vocal inflection changes [8], blood-glucose levels, video recordings of facial expressions and posture gesture changes [9] and other bodily changes [10] [11]. But video recordings make data acquisition expensive and it's also has its own limitations. In the proposed method, features extracted from the biomedical signals in order that they become more robust and reliable.