Driver Drowsiness Detection Using Multi-Channel Second Order Blind Identifications

Abstract

It is well known that blink, yawn, and heart rate changes give clue about a human's mental state, such as drowsiness and fatigue. In this paper, image sequences, as the raw data, are captured from smart phones which serve as non-contact optical sensors. Video streams containing subject's facial region are analyzed to identify the physiological sources that are mixed in each image. We then propose a method to extract blood volume pulse and eye blink and yawn signals as multiple independent sources simultaneously by multi-channel second-order blind identification (SOBI) without any other sophisticated processing, such as eye and mouth localizations. An overall decision is made by analyzing the separated source signals in parallel to determine the driver's driving state. The robustness of the proposed method is tested under various illumination contexts and a variety of head motion modes. Experiments on 15 subjects show that the multi-channel SOBI presents a promising framework to accurately detect drowsiness by merging multi-physiological information in a less complex way. INDEX TERMS Yawn, blink, blood volume pulse (BVP), drowsiness detection, second-order blind identification (SOBI).

A Generic Design of Driver Drowsiness and Stress Recognition Using MOGA Optimized Deep MKL-SVM

“…Typical algorithms for driver drowsiness recognition were based on three types of inputs: (i) the biometric-signal-based approach [16][17][18][19]; (ii) the vehicle-based approach [20][21][22][23] and (iii) the image-based approach [24][25][26][27]. Approach (i) is intrusive whereas approaches (ii) and (iii) are non-intrusive.…”

“…Mandal et al proposed a fusion and reasoning method to measure the driver drowsiness, the following modules were included, head and shoulder detection, face detection based on the front view and oblique view analysis, eye detection based on Open Source Computer Vision Library (OpenCV) and Institute for Infocomm Research (I2R) as well as the eye openness estimation [26]. A multi-channel (3, 6 and 9 channels) second-order blind identification algorithm was proposed, which analyzed the yawn signals and eye blinks and yielded optimal thresholds for the drowsiness level [27].…”

“…Various methods have been proposed for driver drowsiness recognition [16][17][18][19][20][21][22][23][24][25][26][27] and driver stress recognition [28][29][30][31][32][33][34][35]. Nonetheless, there exist several challenges and limitations that require further research.…”

“…First, the comparison was carried out on driver drowsiness recognition. Table 2 summarizes the performance between proposed MOGA optimized D-MKL-SVM and existing works [16][17][18][19][20][21][22][23][24][25][26][27]. It summarizes the category of input signals, dataset, methodology, types of cross-validation and performance.…”

“…Therefore, EEG and video-based approaches experience challenges in signal acquisition. As a result, the performance in related works [18,19,[24][25][26][27] is reduced in a practical situation, which the methodology is governed by a maximum accuracy of 85% and 59% for an EEG-based and image-based approach.…”