This paper presents an effective traffic surveillance system for detecting and tracking moving vehicles in nighttime scenes. The proposed method identifies by locating vehicle headlights taillights using image segmentation pattern analysis techniques. First, a fast bright-object process based on automatic multilevel histogram thresholding is applied to effectively extract bright objects of interest. approach provides robust adaptable detection that operates well under various illumination...

Automated facial expression recognition can greatly improve the human-machine interface. The machine provide better and more personalized services when it knows human's emotion. This kind of improvement is an important progress in this artificial intelligence era. Many deep learning approaches have been applied recent years due to their outstanding accuracy after training with large amounts data. performance limited, however, by specific environmental conditions variations different persons...

In this paper, we propose a high-performance and memory-efficient pipeline architecture which performs the one-level two-dimensional (2-D) discrete wavelet transform (DWT) in 5/3 9/7 filters. general, internal memory size of 2-D highly depends on registers one-dimensional (1-D) DWT. Based lifting-based DWT algorithm, primitive data path is modified an efficient derived to shorten path. Accordingly, under same arithmetic resources, 1-D can operate at higher processing speed (up 200 MHz...

In this paper, an intelligent automated lane-keeping system is proposed and implemented on our vehicle platform, i.e., TAIWAN i TS-1. This challenges the online integrating heterogeneous systems such as a real-time vision system, lateral controller, in-vehicle sensors, steering wheel actuating motor. The detects lane markings ahead of vehicle, regardless varieties in road appearance, determines desired trajectory based relative positions with respect to center road. To achieve more humanlike...

In speech processing, endpoint detection in noisy environments is difficult, especially the presence of nonstationary noise. Robust one most important areas processing. Generally, feature parameters used for are highly sensitive to environment. Endpoint severely degraded at low signal-to-noise ratios (SNRs) since those cannot adequately describe characteristics a signal. As result, this study seeks banded structure on spectrogram distinguish from nonspeech, adverse environments. First,...

Remote photoplethysmography (rPPG) is an unobtrusive solution to heart rate monitoring in drivers. However, disturbances that occur during driving such as driver behavior, motion artifacts, and illuminance variation complicate the of rate. Faced with disturbance, one commonly used assumption periodicity (or spectrum sparsity). Several methods improve stability at expense tracking sensitivity for variation. Based on statistical signal processing (SSP) Monte Carlo simulations, outlier...

This paper proposes a new localization method for mobile robots based on received signal strength (RSS) in indoor wireless local area networks (WLANs). In networks, propagation conditions are very difficult to predict due interference, reflection, and fading effects. As result, an explicit measurement equation is not available. this paper, observation likelihood model accomplished using kernel density estimation characterize the dependence of location RSS. Based measured RSS, robot's...

Abstract Atrial fibrillation (AF) is often asymptomatic and paroxysmal. Screening monitoring are needed especially for people at high risk. This study sought to use camera-based remote photoplethysmography (rPPG) with a deep convolutional neural network (DCNN) learning model AF detection. All participants were classified into groups of AF, normal sinus rhythm (NSR) other abnormality based on 12-lead ECG. They then underwent facial video recording 10 min rPPG signals extracted segmented 30-s...

A simplified approach to W.H. Bode's (1945) theorem for both continuous-time and discrete-time systems, along with some generalization, are presented. For the constraints of open-loop stability roll-off at s= varies as removed. counterexample shows that when excess poles/zeros vanishes, Bode integral drops from infinite finite value gain crosses a critical value. revised result is also developed. The salient feature this no stage either Cauchy's or Poisson invoked; proof relies only on...

In this paper, a vehicle detection approach for complex environments is presented. This paper proposes methods solving problems of in traffic jams and weather conditions such as sunny days, rainy cloudy sunrise time, sunset or nighttime. recent research, there have been many well-known detectors that utilize background extraction to recognize vehicles. these studies, the image needs continuously be updated; otherwise, luminance variation will impact quality. The under various difficulties...

This study presents an effective method for detecting vehicles in front of the camera-assisted car during nighttime driving. The proposed detects based on and locating vehicle headlights taillights using techniques image segmentation pattern analysis. First, to effectively extract bright objects interest, a process automatic multilevel thresholding is applied grabbed road-scene images. Then extracted are processed by rule-based procedure, identify analyzing their light patterns, estimate...

Advanced driver-assistance systems have recently become one of the most active topics related to intelligent vehicles. Such assistance facilitates vehicular operation by allowing drivers increased control and an enhanced driving experience, it even comes in form automated for autonomous vehicle operation. This paper presents a system that uses low-cost embedded digital signal processor, with overall installed commercial vehicle. Based on information supplied multiple sensors, such as...

This paper presents a relative discriminative histogram of oriented gradients (HOG) (RDHOG)-based particle filter (RDHOGPF) approach to traffic surveillance with occlusion handling. Based on the conventional HOG, an extension known as RDHOG is proposed, which enhances descriptive ability central block and surrounding blocks. RDHOGPF can be used predict update positions vehicles in continuous video sequences. was integrated framework order improve tracking robustness accuracy. To resolve...

The driver's physiological status has enormous value to public traffic safety and cannot be ignored nowadays. In addition, heart rate (HR) is one of the most important indicators human's health status. When detecting HR, using traditional contact-type devices might bring about distraction or discomfort. On contrary, remote photoplethysmography (rPPG) technique a better way monitor HR in vehicle applications simply by web-camera without interfering driver. Most rPPG studies intended reduce...

Remote photoplethysmography (rPPG) measurement has received much attention due to its attractive applications in contactless physiological monitoring. It calculates the blood volume pulse from color signals with a normal camera. Artifacts caused by motion and illumination changes must be accounted for, especially real-world usage. In this article, we propose novel framework that includes an error compensation neural network after conventional signal-based rPPG extraction improve measuring...

Atrial fibrillation (AF) has been proven highly correlated to stroke; more than 43 million people suffer from AF worldwide. However, most of these patients are unaware their disease. There is no convenient tool by which conduct a comprehensive screening identify asymptomatic patients. Hence, we provide non-contact detection approach based on remote photoplethysmography (rPPG). We address motion disturbance, the challenging issue in rPPG technology, with NR-Net, ATT-Net, and SQ-Mask modules....

Blood pressure (BP) is a major vital sign that highly correlated with human health. However, for decades BP measurement has involved cuff, which can be uncomfortable and even carry the risk of infection. The remote photoplethysmography (rPPG) technique gained great success in heart rate monitoring tackling these problems. Some studies integrate pulse transit time (PTT) into system ask users put their hand beside face. A calibration procedure usually needed such methods. Other methods measure...

This paper presents a safe and comfortable longitudinal automation system which incorporates human-in-the-loop technology. The proposed has hierarchical structure that consists of an adaptive detection area, supervisory control, regulation control. area routes the information from on-board sensors to ensure vehicles ahead, particularly when driving on curves. Based recognized target distance control determines desired velocity for vehicle maintain safety smooth operation in different modes....

This study presents an effective system for detecting and tracking moving vehicles in nighttime traffic scene surveillance. The proposed method identifies based on locating vehicle headlights taillights by using the techniques of image segmentation pattern analysis. First, to effectively extract bright objects interest, a fast bright-object process automatic multilevel histogram thresholding is applied road-scene images. approach can provide robustness adaptability detection be operated well...

An approach of rapidly computing the projective width lanes is presented to predict positions and widths lanes. The Lane Marking Extraction Finite State Machine designed extract points with features lane markings in image, a cubic B-spline adopted conduct curve fitting reconstruct road geometry. A statistical search algorithm also proposed correctly adaptively determine thresholds under various kinds illumination conditions. Furthermore, parameters camera moving car may change vibration, so...

With the growing concern for driving safety, many driving-assistance systems have been developed. In this paper, we develop a reasoning-based framework monitoring of safety. The main objective is to present drivers with an intuitively understood green/yellow/red indicator their danger level. Because level may change owing interaction host vehicle and environment, proposed involves two stages danger-level alerts. first stage collects lane bias, distance front car, longitudinal lateral...

This system counts passengers through a single camera that is fixed at an overhead (zenithal) position. The proposed algorithm designed to solve the problem of illumination, which can occur when bus door opens or closes. A processing ratio about 30 frames/s necessary overcome extreme change in illumination. combines global positioning data and standard time establish number each stop different periods calculates unoccupied seats provide business-related information.

Deriving blood pressure in a non-invasive way via photoplethysmography (PPG) signals has become familiar topic. With the knowledge of relation between PPG and pressure, we expect to further make measurement contactless for convenience reasons. An alternative signal source is remote (rPPG) signals. There are mainly two kinds approaches exploiting through signals, one by calculating pulse transit time arterial wave at consecutive sites other based on waveform feature analysis from single...