Automatic Incident Detection Algorithm Research Articles

An Automatic Incident Detection Method for a Vehicle-to-Infrastructure Communication Environment: Case Study of Interstate 64 in Missouri

Transportation agencies continuously and consistently work to improve the processes and systems for mitigating the impacts of roadway incidents. Such efforts include utilizing emerging technologies to reduce the detection and response time to roadway incidents. Vehicle-to-infrastructure (V2I) communication is an emerging transportation technology that enables communication between a vehicle and the infrastructure. This paper proposes an algorithm that utilizes V2I probe data to automatically detect roadway incidents. A simulation testbed was developed for a segment of Interstate 64 in St. Louis, Missouri to evaluate the performance of the V2I-based automatic incident detection algorithm. The proposed algorithm was assessed during peak and off-peak periods with various incident durations, under several market penetration rates for V2I technology, and with different spatial resolutions for incident detection. The performance of the proposed algorithm was assessed on the basis of the detection rate, time to detect, detection accuracy, and false alarm rate. The performance measures obtained for the V2I-based automatic incident detection algorithm were compared with California #7 algorithm performance measures. The California #7 algorithm is a traditional automatic incident detection algorithm that utilizes traffic sensors data, such as inductive loop detectors, to identify roadway events. The California #7 algorithm was implemented in the Interstate 64 simulation testbed. The case study results indicated that the proposed V2I-based algorithm outperformed the California #7 algorithm. The detection rate for the proposed V2I-based incident detection algorithm was 100% in market penetrations of 50%, 80%, and 100%. However, the California #7 algorithm's detection rate was 71%.

Advanced Fuzzy-Logic-Based Traffic Incident Detection Algorithm

This study demonstrates an incident detection algorithm that uses the meteorological and traffic parameters for improving the poor performance of the automatic incident detection (AID) algorithms under extreme weather conditions and for efficiently using the meteorological devices on advanced freeways. This algorithm comprises an incident detection module that is based on learning vector quantization (LVQ) and a meteorological influencing factor module. Field data are obtained from the Yuwu freeway in Chongqing, China, to verify the algorithm. Further, the performance of this algorithm is evaluated using commonly used criteria such as mean time to detection (MTTD), false alarm rate (FAR), and detection rate (DR). Initially, an experiment is conducted for selecting the algorithm architecture that yields the optimal detection performance. Additionally, a comparative experiment is performed using the California algorithm, exponential smoothing algorithm, standard normal deviation algorithm, and McMaster algorithm. The experimental results demonstrate that the algorithm proposed in this study is characterized by high DR, low FAR, and considerable suitability for applications in AID.

Extreme Learning Machine-Based Traffic Incidents Detection with Domain Adaptation Transfer Learning

Abstract Traffic incidents in big cities are increasing alongside economic growth, causing traffic delays and deteriorating road safety conditions. Thus, developing a universal freeway automatic incident detection (AID) algorithm is a task that took the interest of researchers. This paper presents a novel automatic traffic incident detection method based on the extreme learning machine (ELM) algorithm. Furthermore, transfer learning has recently gained popularity as it can successfully generalise information across multiple tasks. This paper aimed to develop a new approach for the traffic domain-based domain adaptation. The ELM was used as a classifier for detection, and target domain adaptation transfer ELM (TELM-TDA) was used as a tool to transfer knowledge between environments to benefit from past experiences. The detection performance was evaluated by common criteria including detection rate, false alarm rate, and others. To prove the efficiency of the proposed method, a comparison was first made between back-propagation neural network and ELM; then, another comparison was made between ELM and TELM-TDA.

Automatic traffic incident detection algorithm for both rain and no-rain conditions

This study proposes a flow-rain-dependent algorithm for the automatic detection of traffic incidents under both rain and no-rain conditions. To overcome the shortcomings of discrete detection thresholds in previous automatic incident detection (AID) algorithms, continuous detection thresholds are used. These thresholds are generated by calibrated generalized detection threshold functions, in which both pre-incident traffic and conditions in rain are explicitly modeled. The volume/capacity ratio, which can better describe the degree of congestion, is used to indicate the pre-incident traffic conditions in the proposed algorithm. A case study is carried out on a territory-wide road network in Hong Kong to demonstrate the performance of the proposed AID algorithm. Traffic data for journey time estimation is collected from the Hong Kong urban road network. The results show that incorporating the rain effect when determining the detection threshold can improve the overall performance of traffic incident detection. Language: en

Development of an AIDA(Automatic Incident Detection Algorithm) for Uninterrupted Flow Based on the Concept of Short-term Displaced Flow

Many traffic centers are highly hesitant in employing existing Automatic Incident Detection Algorithms due to high false alarm rate, low detection rate, and enormous effort taken in maintaining algorithm parameters, together with complex algorithm structure and filtering/smoothing process. Concerns grow over the situation particularly in Freeway Incident Management Area This study proposes a new algorithm and introduces a novel concept, the Displaced Flow Index (DiFI) which is similar to a product of relative speed and relative occupancy for every execution period. The algorithm structure is very simple, also easy to understand with minimum parameters, and could use raw data without any additional pre-processing. To evaluate the performance of the DiFI algorithm, validation test on the algorithm has been conducted using detector data taken from Naebu Expressway in Seoul and following transferability tests with Gyeongbu Expressway detector data. Performance test has utilized many indices such as DR, FAR, MTTD (Mean Time To Detect), CR (Classification Rate), CI (Composite Index) and PI (Performance Index). It was found that the DR is up to 100%, the MTTD is a little over 1.0 minutes, and the FAR is as low as 2.99%. This newly designed algorithm seems promising and outperformed SAO and most popular AIDAs such as APID and DELOS, and showed the best performance in every category.

Automatic detection of urban traffic incidents and supporting decision model for police dispatching based on travel time

It is very important to get the complete and timely information of urban road traffic incidents and then to make a reasonable strategy for police dispatching. By improving the efficiency of sending police, the loss of traffic incidents and the pressure of traffic police will be reduced greatly. An assistant decision model of police dispatching based on the information of automatic traffic incident detection is proposed in this paper. Firstly, an automatic traffic incident detection algorithm is put forward based on travel time of urban road section. Two severity dimensions of traffic incidents could be detected, and the average detection time could be reduced significantly. Then, an assistant decision model of police dispatching is established based on Bayesian decision theory by utilizing the results of automatic incident detection algorithm as well as the experience of police officer who in charge of police duty scheduling. Even though the police officer doesn't get the clear and complete information of incidents; he can also qualify the probability of actual states of different traffic incidents with the aided model. Case studies indicate that the model can help police officer to make the strategies of police dispatching and then reduce the risks of decision-making at a certain extent.

Geovisualization of Archived ITS Data-Case Studies

Although transportation data have a spatial component by nature, this component generally has not been fully utilized when it comes to intelligent transportation system (ITS) archived data. This paper demonstrates the use of geographic information system (GIS) to spatially visualize and analyze transportation management center (TMC)-related performance measures derived from archived ITS data from a traffic management center in Texas, namely, San Antonio's TransGuide. The paper discusses three case studies. The first case study involved a performance evaluation of an automatic incident detection algorithm deployed by many traffic management centers after a few years of implementation in the field. The overall detection rate was found to be about 20%, and the overall false rate was about 0.005%. The second case study summarizes work completed to address quality control issues associated with a very large archived ITS data set composed of some 3.4 billion 20-s loop detector data records. There were some 1.6 billion (about 48%) speed, volume, and occupancy records that had a quality control flag. Approximately 1.5 billion flagged records were “valid” records, and the remaining 126 million (about 3.7%) flagged records were “abnormal” records. In the third case study, the researchers evaluated the data completeness both at the aggregate level (by server) and at a more detailed individual detector level. At the server level, the completeness rate varied from 95% to 100%. At the individual detector level, the analysis showed that, on average, the completeness rate for all detectors was about 80%. The researchers utilized GIS to prepare maps showing the spatial distribution of incident detection rate, false alarm rate, quality control flags, and data completeness rate. It is worth mentioning that all of the analysis was performed at a detailed segment-by-segment level, which is by itself a unique addition. The researchers recommend implementing the demonstrated spatial analysis of the archived ITS data. This will allow TMC officials to identify spots with any abnormal behavior, whether it is at the corridor level or even at the segment level.

Resampling Methods for Solving Class Imbalance Problem in Traffic Incident Detection

Traffic data is highly skewed with rare traffic incidents in the real word while most of the existing automatic incident detection (AID) algorithms suffer from limitations due to their inability to detect incidents under imbalanced traffic data condition. This paper developed feasible AID algorithms based on resampling methods to process imbalanced traffic data. In order to obtain the optimal sampling method for incident detection, we compare the detection performance of different AID algorithms based on various resampling methods. The detection performance is evaluated by the common criteria including classification rate, detection rate, false alarm rate, mean time to detection and an integrated performance index. The I-880 dataset is finally used in experiments to verify the proposed algorithms. The experimental results indicate that the proposed AID algorithm based on resampling can achieve better performance through handling imbalanced traffic data problem. Moreover, the under-sampling is competitive than over-sampling for traffic incident detection.

CCTV 영상처리를 이용한 터널 내 사고감지 알고리즘

Almost of current Automatic Incident Detection(AID) algorithms involve the vulnerability that detects the traffic accident in open road or in tunnel as the traffic jam not as the traffic accident. This paper proposes the improved accident detection algorithm to enhance the detection probability based on accident detection algorithms applied in open roads. The improved accident detection algorithm provides the preliminary judgment of potential accident by detecting the stopped object by Gaussian Mixture Model. Afterwards, it measures the detection area is divided into blocks so that the occupancy rate can be determined for each block. All experimental results of applying the new algorithm on a real incident was detected image without error.

Customization of Automatic Incident Detection Algorithms for Signalized Urban Arterials

Non-recurrent congestion or incidents are detrimental to the operability and efficiency of busy urban transport networks. There exist multiple automatic incident detection algorithms (AIDAs) to remotely detect the occurrence of an incident in highway or freeway scenarios; however, very little research has been performed to automatically detect incidents in signalized urban arterials. This limited research attention has mostly been focused on developing new urban arterial specific algorithms, rather than identifying alternative methods to synthesize existing freeway-based algorithms for urban conditions. The main hindrance to such synthesis is that the traffic patterns on the signalized urban arterials are significantly different from the same on highways/freeways due to the presence of traffic intersections. This article introduces a new strategy of customizing the existing AIDAs (freeway based or otherwise) to significantly improve their adaptability to signalized urban arterial transport networks. The new strategy focuses on preprocessing the traffic information before being used as input to a freeway/highway-based AIDA to lessen the effect of traffic signals and to imitate the input patterns in highway/freeway-based incident conditions. The effectiveness of this new strategy has been established with the help of four existing AIDAs. The proposed strategy is a simple solution to implement existing algorithms to signalized urban networks without any further instrumentation or operational cost.

Use Subsampling to Solve Imbalanced Dataset Problem for Automatic Incident Detection Algorithm

Considering the fact that the amount of traffic incident data is rare compared to the large amount of normal traffic state data in the real word, we proposed an Automatic Incident Detection (AID) algorithm based on subsampling method. First, an improved subsampling method based on Edited Nearest Neighbor Rule (ENN) algorithm was used to reconstruct the training set to get a balanced dataset. Then, the Support Vector Machine (SVM) was adopted as a classifier to detect traffic incidents. The real traffic data collected from the I-880 freeway in American was used to build the model and test the performance of the proposed AID algorithm. In addition, we made a comparison of the detection performances between the AID algorithm obtained by the original training set and the one by the relatively balanced training set. The experimental results show that the proposed AID algorithm based on subsampling is suitable for imbalanced dataset and can obtain a better detection performance.

Algorithm fusion method to enhance automatic incident detection on Melbourne freeways

This paper addresses the transferability issue faced by many practitioners in developing an effective and efficient automatic incident detection algorithm for different freeways. An algorithm fusion procedure developed for the Central Expressway in Singapore is evaluated to demonstrate its transferability potential in detecting lane-blocking incidents along freeways in Melbourne, Australia. This study observes that the flow-based algorithm fusion options that use a set of different detection threshold values for various pre-incident traffic flow conditions possess promising transferability potential. They give a reasonably high detection rate of above 80% with false alarm rate levels below 0.2% with mean-time-to-detect values less than 150 seconds. These flow-based algorithm fusion options significantly outperform a model specifically developed for traffic conditions on freeways in Melbourne. In conclusion, this method is capable of providing an alternative to the commonly practiced methods in detecting incidents along different sites.

New Automatic Incident Detection Algorithm Based on Traffic Data Collected for Journey Time Estimation

A new automatic incident detection algorithm based on the available data originally collected for journey time estimation in Hong Kong is proposed in this paper. Instead of installing a greater number of expensive detectors, the proposed algorithm has proved feasible in effective traffic incident detection, with the available data collected by both video traffic detectors and automatic vehicle identification readers. The proposed algorithm extends the previous standard normal deviate algorithm in the aspects of mathematical model, input data, and detection logic. Two new traffic parameters are proposed as indicators of incidents. They are the coefficient of variation of speed at the upstream detector and the correlation coefficient of speeds of two adjacent detectors. Historical traffic and accident data on an urban road in Hong Kong are used for calibration and validation of the proposed algorithm. This proposed algorithm outperforms five existing algorithms based on the available data for journey time estimation in Hong Kong. It is expected that the proposed algorithm could be used for incident detection in cities even when data are collected only for journey time estimation.

Evaluation of Texas Incident Detection Algorithm after Years of Implementation

This paper summarizes a performance evaluation of the automatic incident detection algorithm deployed by many traffic management centers across the State of Texas, USA after a few years of implementation in the field. This analysis is the first step towards improving performance of incident detection which will eventually help optimizing freeway incident management practices at these TMCs. To conduct this analysis, the researchers used archived ITS data, particularly freeway incident alarms and incident scenarios data, as well as a geodatabase of ITS features. Matching alarm and scenario data enabled the determination of two performance measures: incident detection rates and false alarm rates. While many researchers evaluated the performance of their proposed incident detection algorithms, only a few assessed the actual performance of these algorithms after long period of implementation. Although the analysis described in this paper uses data from one jurisdiction (San Antonio, Texas), the methodology is sufficiently generic to enable implementation at other traffic management centers. 

Development of an AIDA(Automatic Incident Detection Algorithm) for Uninterrupted Flow By Diminishing the Random Noise Effect of Traffic Detector Variables

The data quality and measurements along consecutive detector stations can vary much even in the same traffic conditions due to variety in detector types, calibration and maintenance effort, field operation periods, minor geometric changes of roads and so on. These faulty situations often create 10% or more of inherent difference in important traffic measurements between two stations even under stable low flow condition. Low detection rates(DR) and high false alarm rates(FAR) therefore sets in among many popular Automatic Incident Detection Algorithms(AIDA). This research is two-folded and aims mainly to develop a new AIDA for uninterrupted flow. For this purpose, a technique which utilizes a Simple Arithmetic Operation(SAO) of traffic variables is introduced. This SAO technique is designed to address the inherent discrepancy of detector data observed successive stations, and to overcome the degradation of AIDA performance. It was found that this new algorithm improves DR as much as 95 percent and above. And mean time to detection(MTTD) is found to be 1 minutes or less. When it comes to FAR, this new approach compared to existing AIDAs reduces FAR up to 31.0 percent. And capability in persistency check of on-going incidents was found excellent as well.

Development of an AIDA(Automatic Incident Detection Algorithm) for Uninterrupted Flow By Diminishing the Random Noise Effect of Traffic Detector Variables

Development of an AIDA(Automatic Incident Detection Algorithm) for Uninterrupted Flow By Diminishing the Random Noise Effect of Traffic Detector Variables

Automatic Freeway Incident Detection Based on Fundamental Diagrams of Traffic Flow

Algorithms for automatic incident detection (AID) detect traffic incidents on the basis of traffic flow measurements. There are two important steps in an AID algorithm: traffic flow feature generation and incident decision making. In the past decade, the research on freeway AID algorithms has been focused on using artificial intelligence to optimize the decision making of AID algorithms. In this paper, the primary focus is on finding a new set of variables for the feature generation. The new variables, uncongested and congested regime shifts (URS and CRS), are generated by conducting coordinate transformation on loop-detected flow and occupancy measurements. A novel AID algorithm, the fundamental diagrams–based automatic incident detection (FD AID) algorithm, is then developed by implementing the incident-related traffic flow knowledge using those variables. Preliminary results show superior performance of the FD algorithm compared with legacy algorithms.

Algorithm Fusion for Detecting Incidents on Singapore’s Central Expressway

Many studies have focused mainly on the development of a single automatic incident detection algorithm to detect incident occurrences, and the performance obtained has often been mixed. In this study, an algorithm fusion method was developed using incident data collected from the Central Expressway in Singapore. This method explores the possibility of enhancing incident detection performance by combining the complementary advantages of a group of existing algorithms. It was found that the fused algorithm with the following options has outperformed the existing algorithms: Fusion Option II—selecting a combination of the best-performing algorithms, and Fusion Option III—applying different weightings to a group of algorithms. Compared with the existing dual variable/combined detector evaluation algorithms developed earlier for the same studied site, the fused algorithms performed significantly better, with false alarm rates between 0.2 and 1.0%. At a detection rate of 90%, these fused algorithms were able to...

Heavy Flow-Based Incident Detection Algorithm Using Information From Two Adjacent Detector Stations

In heavy traffic flow conditions, vehicles have limited manouverability which affects the magnitude of response to incident-induced traffic disturbances and how fast changes in these traffic variables can signal the occurrence of an incident. Such characteristics are usually used to formulate a loop-based algorithm. A recent study reported that some existing algorithms were not able to maintain a desired level of effectiveness when these algorithms were used to detect incidents with a video-based detector system. Two new video-based automatic incident detection algorithms, the INdividual Detection Evaluation (INDE) and COmbined Detection Evaluation (CODE) algorithms were, therefore, developed for the detection of lane-blocking incidents in heavy traffic flow conditions using the Central Expressway in Singapore as a case study. The algorithms detect incident-induced traffic speed and occupancy disturbances differently: INDE processes information at each individual detector station and CODE processes information at two adjacent detector stations. Both algorithms outperformed existing algorithms commonly used in incident management systems. Of these algorithms, the INDE algorithm raised fewer false alarms and gave slightly faster incident warnings. However, its performance was relatively less consistent when applied to a validation database.

Single-Station Algorithm Using Video-Based Data for Detecting Expressway Incidents

: Most automatic incident detection algorithms were successfully developed using loop-detector-based traffic measurements collected from their own localities. But their detection performances were not satisfactory when applied on data collected using a video-based detector system. The video-based detector system is gaining popularity as it was reported to be cost-effective, less prone to damage compared to loop detectors embedded in road pavement, and possesses surveillance capability. It is able to provide the homogeneity of traffic measurements with greater reliability in non-incident situations. In this study, a simple detection rule was used to develop algorithms that use video-based data for detecting lane-blocking incidents. A set of 96 incidents from Singapore's Central Expressway was used for calibrating these algorithms, with another 64 incidents for validation. Two single-station algorithms, named dual-variable (DV) and flow-based DV algorithms were developed. They have similar detection logic, but the latter includes a pre-incident traffic flow condition in its detection framework. On average, the flow-based DV algorithm outperformed the DV algorithm, and both proved to be effective techniques when compared to some existing loop-detector-based algorithms.