Automated vision inspection of rail surface cracks: A double-layer data-driven framework

A double-layer data-driven framework for the automated vision inspection of the rail surface cracks is proposed in this paper. Based on images of rails, the proposed framework is capable to detect the location of cracks firstly and next automatically obtain the boundary of cracks via a feature-based linear iterative crack aggregation (FLICA). Extended Haar-like features are applied to develop significant features for identifying cracks in images. Built on extended Haar-like features, a cascading classifier ensemble integrating three single cascading classifiers with a major voting scheme is proposed to decide the presence of cracks in the image. Each single cascading classifier is composed of a sequence of stage classifiers trained by the LogitBoost algorithm. A scalable sliding window carrying the cascading classifier ensemble is applied to scan images of rail tracks, which is identified by the Otsu’s method, and detect cracks. After completing the crack registration in the first layer, the FLICA is developed to discover boundaries of cracks. The effectiveness of the proposed data-driven framework for identifying rail surface cracks is validated with the rail images provided by the China Railway Corporation and Hong Kong Mass Transit Railway (MRT). Six benchmarking methods, the Otsu’s method, mean shift, the visual detection system, the geometrical approach, fully convolutional networks and the U-net, are utilized to prove advantages of the proposed framework. Results of the validation and comparative analyses demonstrate that the proposed framework is most effective in the rail surface crack detection.     

Using automated walking gait analysis for the identification of pedestrian attributes

Collecting microscopic pedestrian behavior and characteristics data is important for optimizing the design of pedestrian facilities for safety, efficiency, and comfortability. This paper provides a framework for the automated classification of pedestrian attributes such as age and gender based on information extracted from their walking gait behavior. The framework extends earlier work on the automated analysis of gait parameters to include analysis of the gait acceleration data which can enable the quantification of the variability, rhythmic pattern and stability of pedestrian’s gait. In this framework, computer vision techniques are used for the automatic detection and tracking of pedestrians in an open environment resulting in pedestrian trajectories and the speed and acceleration dynamic profiles. A collection of gait features are then derived from those dynamic profiles and used for the classification of pedestrian attributes. The gait features include conventional gait parameters such as gait length and frequency and dynamic parameters related to gait variations and stability measures. Two different techniques are used for the classification: a supervised k-Nearest Neighbors (k-NN) algorithm and a newly developed semi-supervised spectral clustering. The classification framework is demonstrated with two case studies from Vancouver, British Columbia and Oakland, California. The results show the superiority of features sets including gait variations and stability measures over features relying only on conventional gait parameters. For gender, correct classification rates (CCR) of 80% and 94% were achieved for the Vancouver and Oakland case studies, respectively. The classification accuracy for gender was higher in the Oakland case which only considered pedestrians walking alone. Pedestrian age classification resulted in a CCR of 90% for the Oakland case study.     

A hybrid Bayesian Network approach to detect driver cognitive distraction

Driver cognitive distraction (e.g., hand-free cell phone conversation) can lead to unapparent, but detrimental, impairment to driving safety. Detecting cognitive distraction represents an important function for driver distraction mitigation systems. We developed a layered algorithm that integrated two data mining methods—Dynamic Bayesian Network (DBN) and supervised clustering—to detect cognitive distraction using eye movement and driving performance measures. In this study, the algorithm was trained and tested with the data collected in a simulator-based study, where drivers drove either with or without an auditory secondary task. We calculated 19 distraction indicators and defined cognitive distraction using the experimental condition (i.e., “distraction” as in the drives with the secondary task, and “no distraction” as in the drives without the secondary task). We compared the layered algorithm with previously developed DBN and Support Vector Machine (SVM) algorithms. The results showed that the layered algorithm achieved comparable prediction performance as the two alternatives. Nonetheless, the layered algorithm shortened training and prediction time compared to the original DBN because supervised clustering improved computational efficiency by reducing the number of inputs for DBNs. Moreover, the supervised clustering of the layered algorithm revealed rich information on the relationship between driver cognitive state and performance. This study demonstrates that the layered algorithm can capitalize on the best attributes of component data mining methods and can identify human cognitive state efficiently. The study also shows the value in considering the supervised clustering method as an approach to feature reduction in data mining applications.     

Classifiers for driver activity monitoring

The goal of this work is the detection and classification of driver activities in an automobile using computer vision. To this end, this paper presents a novel two-step classification algorithm, namely, an unsupervised clustering algorithm for grouping the actions of a driver during a certain period of time, followed by a supervised activity classification algorithm. The main contribution of this work is the combination of the two methods to provide a computationally fast solution for deployment in real-world scenarios that is robust to illumination and segmentation issues under most conditions experienced in the automobile environment. The unsupervised clustering groups the actions of the driver based on the relative motion detected using a skin-color segmentation algorithm, while the activity classifier is a binary Bayesian eigenimage classifier. Activities are grouped as safe or unsafe and the results of the classification are shown on several subjects obtained from two distinct driving video sequences.     

Automated classification based on video data at intersections with heavy pedestrian and bicycle traffic: Methodology and application

Pedestrians and cyclists are amongst the most vulnerable road users. Pedestrian and cyclist collisions involving motor-vehicles result in high injury and fatality rates for these two modes. Data for pedestrian and cyclist activity at intersections such as volumes, speeds, and space–time trajectories are essential in the field of transportation in general, and road safety in particular. However, automated data collection for these two road user types remains a challenge. Due to the constant change of orientation and appearance of pedestrians and cyclists, detecting and tracking them using video sensors is a difficult task. This is perhaps one of the main reasons why automated data collection methods are more advanced for motorized traffic. This paper presents a method based on Histogram of Oriented Gradients to extract features of an image box containing the tracked object and Support Vector Machine to classify moving objects in crowded traffic scenes. Moving objects are classified into three categories: pedestrians, cyclists, and motor vehicles. The proposed methodology is composed of three steps: (i) detecting and tracking each moving object in video data, (ii) classifying each object according to its appearance in each frame, and (iii) computing the probability of belonging to each class based on both object appearance and speed. For the last step, Bayes’ rule is used to fuse appearance and speed in order to predict the object class. Using video datasets collected in different intersections, the methodology was built and tested. The developed methodology achieved an overall classification accuracy of greater than 88%. However, the classification accuracy varies across modes and is highest for vehicles and lower for pedestrians and cyclists. The applicability of the proposed methodology is illustrated using a simple case study to analyze cyclist–vehicle conflicts at intersections with and without bicycle facilities.     

基于视频的行人检测算法研究

文章提出了一种基于视频图像的行人检测算法。通过动态阈值目标分割算法获得运动目标,同时将分割结果块化处理,以减少算法时间复杂度。使用最小外接矩形获得目标的高宽比,结合基于质心的目标跟踪方法获得目标物体运动速度,进而对行人事件进行检测。实验结果表明,该方法能准确地检测交通场景中的行人事件,有效地解决抖动、光线及积水等因素对行人目标检测的影响,具有良好的实时性和准确性。     

Categorizing bicycling environments using GPS-based public bicycle speed data

A promising alternative transportation mode to address growing transportation and environmental issues is bicycle transportation, which is human-powered and emission-free. To increase the use of bicycles, it is fundamental to provide bicycle-friendly environments. The scientific assessment of a bicyclist’s perception of roadway environment, safety and comfort is of great interest. This study developed a methodology for categorizing bicycling environments defined by the bicyclist’s perceived level of safety and comfort. Second-by-second bicycle speed data were collected using global positioning systems (GPS) on public bicycles. A set of features representing the level of bicycling environments was extracted from the GPS-based bicycle speed and acceleration data. These data were used as inputs for the proposed categorization algorithm. A support vector machine (SVM), which is a well-known heuristic classifier, was adopted in this study. A promising rate of 81.6% for correct classification demonstrated the technical feasibility of the proposed algorithm. In addition, a framework for bicycle traffic monitoring based on data and outcomes derived from this study was discussed, which is a novel feature for traffic surveillance and monitoring.     

Optimization of pedestrian phase patterns at signalized intersections: a multi‐objective approach

This paper presents a multi‐objective optimization model and its solution algorithm for optimization of pedestrian phase patterns, including the exclusive pedestrian phase (EPP) and the conventional two‐way crossing (TWC) at an intersection. The proposed model will determine the optimal pedestrian phase pattern and the corresponding signal timings at an intersection to best accommodate both vehicular traffic and pedestrian movements. The proposed model is unique with respect to the following three critical features: (1) proposing an unbiased performance index for comparison of EPP and TWC by explicitly modeling the pedestrian delay under the control of TWC and EPP; (2) developing a multi‐objective model to maximize the utilization of the available green time by vehicular traffic and pedestrian under both EPP or TWC; and (3) designing a genetic algorithm based heuristic algorithm to solve the model. Case study and sensitivity analysis results have shown the promising property of the proposed model to assist traffic practitioners, researchers, and authorities in properly selecting pedestrian phase patterns at signalized intersections. Copyright © 2013 John Wiley & Sons, Ltd.     

Computer vision approach for the classification of bike type (motorized versus non‐motorized) during busy traffic in the city of Shanghai

This article describes a novel approach for the binary classification of two‐wheeler road users in a dense mixed traffic intersection. The classification is a supervised procedure to differentiate between motorized and non‐motorized (human‐powered) bikes. Road users were first detected and tracked using object recognition methods. Classification features were then selected from the collected trajectories. The features include maximum speed, cadence frequency in addition to acceleration‐based parameters. Experiments were conducted on a video data set from Shanghai, China, where cyclists as well as motorcycles tend to share the main road facilities. A sensitivity analysis was performed to assess the quality of the selected features in improving the accuracy of the classification. A performance analysis demonstrated the robustness of the proposed classification method with a correct classification rate of up to 93%. This research contributes to the literature of automated data collection and can benefit the applications in many transportation‐related fields such as shared space facility planning, simulation models for two‐wheelers, and behavior analysis and road safety studies. Copyright © 2015 John Wiley & Sons, Ltd.     

A study of tour-based mode choice based on a Support Vector Machine classifier

A new approach in recognizing travel mode choice patterns is proposed, based on the Support Vector Machine classification technique. The tour-based travel demand dataset that is analysed is for New York State, derived from the 2009 U.S. National Household Travel Survey. The main features characterizing each tour are the means used, travel-related variables and socioeconomic aspects. Results obtained demonstrate the ability to predict to some extent, in real settings where car use dominates, which tours are likely to be made by public transport or non-motorized means. Moreover, the flexibility of the technique allows assessing the predictive power of each feature according to the combination of travel means used in different tours. Potential applications range from activity-based travel choice simulators to search engines supporting personalized travel planners – in general, whenever ‘best guesses’ on mode choice patterns have to be made quickly on large amounts of data prejudicing the possibility of setting up a statistical model.     

Optimization of mid-block pedestrian crossing network with discrete demands

In many cases, pedestrian crossing demands are distributed discretely along an arterial segment. Demand origins, destinations and crosswalks comprise a pedestrian crossing network. An integrated model for optimizing the quantity, locations and signal settings of mid-block crosswalks simultaneously is proposed to best trade-off the operational performances between pedestrians and vehicles. Pedestrian behavior of choosing crosswalks is captured under a discrete demand distribution. Detour distance and delay at signalized crosswalks are formulated as a measure of pedestrian crossing cost. Maximum bandwidths are modeled in analytical expressions as a measure of vehicular cost. To solve the proposed model, the Non-dominated Sorting Genetic Algorithm II (NSGA II) based algorithm is designed and employed to obtain the Pareto frontier efficiently. From the numerical study, it is found that there exists an optimal number of mid-block crosswalks. Excess available crosswalks may make no contributions to improvement in pedestrian cost when the constraint of the minimum interval between crosswalks and vehicular cost are taken into account. Two-stage crosswalks are more favorable than one-stage ones for the benefits of both pedestrian and vehicles. The study results show promising properties of the proposed method to assist transportation engineers in properly designing mid-block crosswalks along a road segment.     

Microscopic pedestrian simulation model combined with a tactical model for route choice behaviour

This study proposes a microscopic pedestrian simulation model for evaluating pedestrian flow. Recently, several pedestrian models have been proposed to evaluate pedestrian flow in crowded situations for the purpose of designing facilities. However, current pedestrian simulation models do not explain the negotiation process of collision avoidance between pedestrians, which can be important for representing pedestrian behaviour in congested situations. This study builds a microscopic model of pedestrian behaviour using a two-player game and assuming that pedestrians anticipate movements of other pedestrians so as to avoid colliding with them. A macroscopic tactical model is also proposed to determine a macroscopic path to a given destination. The results of the simulation model are compared with experimental data and observed data in a railway station. Several characteristics of pedestrian flows such as traffic volume and travel time in multidirectional flows, temporal–spatial collision avoidance behaviour and density distribution in the railway station are reproduced in the simulation.     

公路隧道运营通风横通道周围流场特性研究

为研究公路隧道巷道式运营通风横通道周围的流场特性,以贵州凉风坳隧道为研究背景,利用流体力学研究软件Fluent建立隧道三维模型并进行了数值仿真计算。通过在隧道两边等风速条件下,对隧道内人行横通道以及车行横通道附近的流场特点研究,分析了运营通风横通道周围流场分布的具体特点。模拟结果表明:车行横通道与人行横通道两边流场分布基本相同;横通道内风速相对较小,约为隧道内风速的10%;不同风速下,横通道附近的速度分布和总压分布整体走势一致;两边流场相对稳定,不会进行风流和污染物的串流。     

A model of pedestrians’ intended waiting times for street crossings at signalized intersections

For the purposes of both traffic-light control and the design of roadway layouts, it is important to understand pedestrian street-crossing behavior because it is not only crucial for improving pedestrian safety but also helps to optimize vehicle flow. This paper explores the mechanism of pedestrian street crossings during the red-man phase of traffic light signals and proposes a model for pedestrians’ waiting times at signalized intersections. We start from a simplified scenario for a particular pedestrian under specific traffic conditions. Then we take into account the interaction between vehicles and pedestrians via statistical unconditioning. We show that this in general leads to a U-shaped distribution of the pedestrians’ intended waiting time. This U-shaped distribution characterizes the nature of pedestrian street-crossing behavior, showing that in general there are a large proportion of pedestrians who cross the street immediately after arriving at the crossing point, and a large proportion of pedestrians who are willing to wait for the entire red-man phase. The U-shaped distribution is shown to reduce to a J-shaped or L-shaped distribution for certain traffic scenarios. The proposed statistical model was applied to analyze real field data.     

A new method for evaluation of level of service in pedestrian facilities

Safe and comfortable walking is essential for pedestrian movement in modern urban transportation systems. Since pedestrian traffic cannot be restricted in some specified streets, some measures for pedestrians have to be taken everywhere in urban areas. This research describes a way to evaluate ordinary sidewalks, and two different methods are proposed. One is an evaluation based on pedestrian behaviour and the other is an evaluation based on pedestrian opinion. Using the indices of pedestrian density and sidewalk width, we can estimate the level of service of sidewalk usage. But generally speaking, since it is not often that a sidewalk is insufficient to deal with pedestrian flow, another approach is necessary for its evaluation, that is, pedestrian awareness of sidewalks must be taken into account. The former method is recommended for all sidewalks, especially with comparatively heavy pedestrian traffic, and the latter method is recommended for ones with light pedestrian traffic.     

A dynamic evacuation model for pedestrian–vehicle mixed-flow networks

In urban emergency evacuation, a potentially large number of evacuees may depend either on transit or other modes, or need to walk a long distance, to access their passenger cars. In the process of approaching the designated pick-up points or parking areas for evacuation, the massive number of pedestrians may cause tremendous burden to vehicles in the roadway network. Responsible agencies often need to contend with congestion incurred by massive vehicles emanating from parking garages, evacuation buses generated from bus stops, and the conflicts between evacuees and vehicles at intersections. Hence, an effective plan for such evacuation needs to concurrently address both the multi-modal traffic route assignment and the optimization of network signal controls for mixed traffic flows. This paper presents an integrated model to produce the optimal distribution of vehicle and pedestrian flows, and the responsive network signal plan for massive mixed pedestrian–vehicle flows within the evacuation zone. The proposed model features its effectiveness in accounting for multiple types of evacuation vehicles, the interdependent relations between pedestrian and vehicle flows via some conversion locations, and the inevitable conflicts between intersection turning vehicle and pedestrian flows. An illustrating example concerning an evacuation around the M&T stadium area has been presented, and the results indicate the promising properties of our proposed model, especially on reflecting the complex interactions between vehicle and pedestrian flows and the favorable use of high-occupancy vehicles for evacuation operations.     

基于SVM的隧道衬砌空洞填充物雷达图像识别研究北大核心CSCD

衬砌背后空洞及其填充物对隧道结构安全具有重要影响,开展空洞探测识别对于结构安全评估和病害处置具有重要意义。首先采用室内试验和FDTD正演模拟相结合的方法,获得了空洞内填充空气、水、干砂、湿砂条件下的雷达图谱数据,并对不同填充物波形规律进行对比分析;然后,基于支持向量机算法对波形特征进行提取和分类识别,建立了一种空洞填充物的人工智能辨识方法。研究结果表明,采用傅里叶变换前的平均值、方差、平均绝对离差和傅里叶变换后的最大幅度值max(fft(X))四个统计量作为支持向量机的识别特征,可以有效区分出衬砌背后填充物的六种类型;当采取单一倾向数据时,识别准确率较好,六种物质二分类问题准确率均可以达到90%以上。     

Improved waveform‐feature‐based vehicle classification using a single‐point magnetic sensor

Vehicle classification systems have important roles in applications related to real‐time traffic management. They also provide essential data and necessary information for traffic planning, pavement design, and maintenance. Among various classification techniques, the length‐based classification technique is widely used at present. However, the undesirable speed estimates provided by conventional data aggregation make it impossible to collect reliable length data from a single‐point sensor during real‐time operations. In this paper, an innovative approach of vehicle classification will be proposed, which achieved very satisfactory results on a single‐point sensor. This method has two essential parts. The first concerns with the procedure of smart feature extraction and selection according to the proposed filter–filter–wrapper model. The model of filter–filter–wrapper is adopted to make an evaluation on the extracted feature subsets. Meanwhile, the model will determine a nonredundant feature subset, which can make a complete reflection on the differences of various types of vehicles. In the second part, an algorithm for vehicle classification according to the theoretical basis of clustering support vector machines (C‐SVMs) was established with the selected optimal feature subset. The paper also uses particle swarm optimization (PSO), with the purpose of searching for an optimal kernel parameter and the slack penalty parameter in C‐SVMs. A total of 460 samples were tested through cross validation, and the result turned out that the classification accuracy was over 99%. In summary, the test results demonstrated that our vehicle classification method could enhance the efficiency of machine‐learning‐based data mining and the accuracy of vehicle classification. Copyright © 2014 John Wiley & Sons, Ltd.     

A dynamic Bayesian network model for real-time crash prediction using traffic speed conditions data

Traffic crashes occurring on freeways/expressways are considered to relate closely to previous traffic conditions, which are time-varying. Meanwhile, most studies use volume/occupancy/speed parameters to predict the likelihood of crashes, which are invalid for roads where the traffic conditions are estimated using speed data extracted from sampled floating cars or smart phones. Therefore, a dynamic Bayesian network (DBN) model of time sequence traffic data has been proposed to investigate the relationship between crash occurrence and dynamic speed condition data. Moreover, the traffic conditions near the crash site were identified as several state combinations according to the level of congestion and included in the DBN model. Based on 551 crashes and corresponding speed information collected on expressways in Shanghai, China, DBN models were built with time series speed condition data and different state combinations. A comparative analysis of the DBN model using flow detector data and a static Bayesian network model was also conducted. The results show that, with only speed condition data and nine traffic state combinations, the DBN model can achieve a crash prediction accuracy of 76.4% with a false alarm rate of 23.7%. In addition, the results of transferability testing imply that the DBN models are applicable to other similar expressways with 67.0% crash prediction accuracy.