The European Commission (EC) has published a European Union (EU) Road Safety Framework for the period 2021 to 2030 to reduce road fatalities. In addition, the EC with the EU Directive 2019/1936 ...requires a much more detailed recording of road attributes. Therefore, automatic detection of school routes, four classes of crosswalks, and divided carriageways were performed in this paper. The study integrated satellite imagery as a data source and the Yolo object detector. The satellite Pleiades Neo 3 with a spatial resolution of 0.3 m was used as the source for the satellite images. In addition, the study was divided into three phases: vector processing, satellite imagery processing, and training and evaluation of the You Only Look Once (Yolo) object detector. The training process was performed on 1951 images with 2515 samples, while the evaluation was performed on 651 images with 862 samples. For school zones and divided carriageways, this study achieved accuracies of 0.988 and 0.950, respectively. For crosswalks, this study also achieved similar or better results than similar work, with accuracies ranging from 0.957 to 0.988. The study also provided the standard performance measure for object recognition, mean average precision (mAP), as well as the values for the confusion matrix, precision, recall, and f1 score for each class as benchmark values for future studies.
The United Nations (UN) stated that all new roads and 75% of travel time on roads must be 3+ star standard by 2030. The number of stars is determined by the International Road Assessment Program ...(iRAP) star rating module. It is based on 64 attributes for each road. In this paper, a framework for highly accurate and fully automatic determination of two attributes is proposed: roadside severity-object and roadside severity-distance. The framework integrates mobile Lidar point clouds with deep learning-based object detection on road cross-section images. The You Only Look Once (YOLO) network was used for object detection. Lidar data were collected by vehicle-mounted mobile Lidar for all Croatian highways. Point clouds were collected in .las format and cropped to 10 m-long segments align vehicle path. To determine both attributes, it was necessary to detect the road with high accuracy, then roadside severity-distance was determined with respect to the edge of the detected road. Each segment is finally classified into one of 13 roadside severity object classes and one of four roadside severity-distance classes. The overall accuracy of the roadside severity-object classification is 85.1%, while for the distance attribute it is 85.6%. The best average precision is achieved for safety barrier concrete class (0.98), while the worst AP is achieved for rockface class (0.72).
Unmanned Aerial Vehicles (UAVs) represent easy, affordable, and simple solutions for many tasks, including the collection of traffic data. The main aim of this study is to propose a new, low-cost ...framework for the determination of highly accurate traffic flow parameters. The proposed framework consists of four segments: terrain survey, image processing, vehicle detection, and collection of traffic flow parameters. The testing phase of the framework was done on the Zagreb bypass motorway. A significant part of this study is the integration of the state-of-the-art pre-trained Faster Region-based Convolutional Neural Network (Faster R-CNN) for vehicle detection. Moreover, the study includes detailed explanations about vehicle speed estimation based on the calculation of the Mean Absolute Percentage Error (MAPE). Faster R-CNN was pre-trained on Common Objects in COntext (COCO) images dataset, fine-tuned on 160 images, and tested on 40 images. A dual-frequency Global Navigation Satellite System (GNSS) receiver was used for the determination of spatial resolution. This approach to data collection enables extraction of trajectories for an individual vehicle, which consequently provides a method for microscopic traffic flow parameters in detail analysis. As an example, the trajectories of two vehicles were extracted and the comparison of the driver’s behavior was given by speed—time, speed—space, and space—time diagrams.
Since the inception of the traffic flow theory, numerous traffic flow models have been formulated by scholars in an effort to more accurately delineate the relationships between various traffic flow ...parameters. However, only a limited number of studies have explored the distinctions between fundamental traffic diagrams, which characterize continuous and interrupted traffic flow conditions. Addressing this research lacuna, we compared twelve “speed–density” and “flow–density” models fitted to empirical data collected under continuous and interrupted traffic flow conditions on a selected regional road in Croatia. The empirical data used to develop these models were extracted from video footage captured by an unmanned aerial vehicle on two representative road segments during characteristic peak and off-peak hours on workdays. Our analysis reveals that, depending on the selected traffic flow model and prevailing traffic flow conditions, the practical capacity of the observed regional road is estimated to be in the range from 799 to 2333 veh/h/lane. It was also discovered that the considered models reach practical capacity at a significantly different density under continuous and interrupted traffic stream conditions, i.e., between 37 and 129 veh/km/lane. The conducted t-tests underscore the need to employ distinct “speed–density” and “flow–density” regression functions for modeling continuous and interrupted traffic stream conditions.
Vehicle speed is one of the main factors that influence the occurrence and severity of the consequences of road traffic accidents. Operating speed can be defined, among other things, as the actual ...speed at which the largest number of road users drive in conditions of free traffic flow. It can be measured on existing roads, however, on newly designed roads it can only be predicted. For this reason, many researchers have examined the correlation between the elements of the road as well as its surroundings and operating speed. By determining the correlation, models for predicting operating speed were created. As part of this paper, the most significant models for predicting operating speed were analysed. Of course, the largest number of models are stochastic, but in recent years, models based on artificial intelligence, more precisely on deep learning, have also been created. Accordingly, the goal of this paper is to review the model for predicting the operating speed of vehicles while identifying opportunities for further research and improvement in this area.
The aim of this paper is to develop a model for forecasting RTA fatalities in Yemen. The yearly fatalities was modeled as the dependent variable, while the number of independent variables included ...the population, number of vehicles, GNP, GDP and Real GDP per capita. It was determined that all these variables are highly correlated with the correlation coefficient (r ≈ 0.9); in order to avoid multicollinearity in the model, a single variable with the highest r value was selected (real GDP per capita). A simple regression model was developed; the model was very good (R2=0.916); however, the residuals were serially correlated. The Prais-Winsten procedure was used to overcome this violation of the regression assumption. The data for a 20-year period from 1991-2010 were analyzed to build the model; the model was validated by using data for the years 2011-2013; the historical fit for the period 1991 - 2011 was very good. Also, the validation for 2011-2013 proved accurate.
Speed is one of the main causes of severe traffic accidents, especially of those involving vulnerable road users. Speed display radars are one of the engineering solutions that aim to reduce the ...speed of traffic flow at locations where it is crucial. A before and after study was conducted at six locations on district roads in the Republic of Croatia. As part of the research, vehicles’ speeds were measured for two weeks, during which the devices’ displays were switched off for one week and switched on for another, however, the devices’ radars recorded and collected the speeds of all vehicles which passed during the research period. The research collected a total of 182,352 speeds of recorded vehicles. The test was performed to determine the effect of speed display radar on the potential vehicles’ speed reduction. The results did not show a statistically significant difference in average hourly speeds at most locations, however, they showed a statistically significant difference in 85th percentile hourly speeds of approximately 1 to 3 km/h. Therefore, it can be concluded that the speed display radars at the locations had a positive effect on reducing the vehicles’ speeds. Additionally, at some locations a decrease in the number of speed limit violators was found which proved to be a statistically significant difference between the share of speed limit violators in total traffic flow.
Road safety is a crucial aspect of transportation systems around the globe, with significant implications for public health economics, and overall quality of life. Reactive road safety assessment ...methods, which focus on identifying and rectifying hazardous locations within road networks, play a key role in maintaining and improving the safety of these systems. The purpose of this paper is to delve into the methodologies currently in use for reactive road safety assessment, with a particular focus on practices in Croatia. By comparing these methods with international approaches, the study aims to not only illuminate the strengths and weaknesses of existing practices, but also to provide valuable recommendations for enhancing road safety assessment in Croatia. Through a detailed analysis of safety performance metrics, network screening methods, and specific methodological elements, the paper offers insights into this critical area of transportation safety.
This article addresses the possibility of improving the traditional bus passenger demand forecasting models by leveraging additional data from relevant big data systems and proposes a conceptual ...framework for developing big data-based forecasting models. Based on the data extracted from available big data systems, the authors have developed a conceptual procedural framework for determining the significance of statistical indicators that can potentially be used as predictor variables for forecasting future passenger demand. At the first stage of the proposed framework, the statistical significance of partial linear correlations between observed statistical indicators and bus ridership demand are determined. All statistical indicators identified as potentially significant are further tested for multicollinearity, homoscedasticity, autocorrelation and multivariate normality to determine the suitability of their inclusion in the final equation of the prediction model. The final formulation of the predictive model was developed using stepwise regression. The R programming language was used to implement the proposed procedural framework to develop a model suitable for predicting passenger demand on the Prizren-Zagreb international bus route. Two predictor variables identified as the most statistically significant are the population of Kosovo and the annual number of Kosovo citizens crossing the Croatian border by bus.
•A generic model for evaluating traffic accident locations is proposed.•The potential use of OpenStreetMap in traffic accident analysis is evaluated.•Traffic accidents street names are analyzed using ...the Jaro–Winkler metrics.•Traffic accident locations are analyzed using Inverse Distance Weighting.•The proposed model shows significant improvements in estimating correct locations.
The objective of this research was to develop a model for validating traffic accident locations that would be applicable worldwide, regardless of linguistic or cultural differences. In order to achieve this, a Volunteered Geographic Information (VGI) dataset was used, the OpenStreetMap (OSM) project. To test the developed model, a total of 8550 accidents with fatal or non-fatal injuries that occurred in the City of Zagreb from 2010 to 2014 were evaluated. Traffic accident data was collected using the pen-and-paper method while the traffic accident locations were determined using Global Positioning System (GPS) receivers embedded within police vehicles. This form of data entry invariably introduces errors in both geometric and contextual attributes. To fully counteract these errors, the developed model consists of two key concepts: the Jaro–Winkler string matching technique and the Inverse Distance Weighting method. Over 66% of traffic accident locations were validated, which is an increase of 15% when compared to the classical approach. The model outlined in this paper shows a significant improvement in estimating the correct location of traffic accidents. This in turn results in a drastic decrease in resources needed to estimate the quality of accident locations.