Anhui Keli Information Industry Co Ltd
Anhui Keli Information Industry Co Ltd
Lu Q.,Hefei University of Technology |
Tao G.,Anhui Keli Information Industry Co. |
Chen Y.,Hefei University of Technology
Multimedia Tools and Applications | Year: 2017
Image retargeting is a process to change the resolution of image while preserve interesting regions and avoid obvious visual distortion. In other words, it focuses on image content more than anything else that applies to filter the useful information for data analysis. Existing approaches may encounter difficulties on the various types of images since most of these approaches only consider 2D features, which are sensitive to the complexity of the contents in images. Researchers are now focusing on the RGB-D information, hoping depth information can help to promote the accuracy. However it is not easy to obtain the RGB-D image we need anywhere and how to utilize depth information is still at the exploration stage. In this paper, instead of using RGB-D data captured by 3D camera, we employ an iterative MRF learning model to predict depth information from a single still image. Then we propose our self-learning 3D saliency model based on the RGB-D data and apply it on the seam carving framework. In seam caving, the self-learning 3D saliency is combined with L1-norm of gradient for better seam searching. Experimental results demonstrate the advantages of our method using RGB-D data in the seam carving framework. © 2017 Springer Science+Business Media New York
Chen Y.,Hefei University of Technology |
Tao G.,Anhui Keli Information Industry Co. |
Ren H.,Hefei University of Technology |
Lin X.,Hefei University of Technology |
Zhang L.,Hefei University of Technology
Neurocomputing | Year: 2016
Seat belt detection in intelligent transportation systems is an important research area, but the current algorithms for such systems are not very well developed. Existing methods are mostly based on edge detection and the Hough transform. However, there are many kinds of vehicles and background environments, which produce many possible edges; thus, these methods often produce false positives. We therefore propose a seat belt detection algorithm for complex road backgrounds based on multi-scale feature extraction using deep learning. We first extract multi-scale features from the regions of the labeled vehicle, windshield, and seat belt to train the detection models using convolution neural network (CNN). Then the coarse candidates of the vehicle, windshield, and seat belt in the test image are detected. For the accurate detection results, a post-processing is employed by using the detection scores as well as the relative positions of these vehicle components to train a classification model through support vector machine (SVM). Finally, we perform a fine mapping and identification process using this classification model on the seat belt region. This method performed well when applied to a database of images collected by road surveillance cameras. © 2017.
Shao H.-P.,Chang'an University |
Shao H.-P.,Anhui Keli Information Industry Co Ltd |
Yang X.-F.,Liaoning Province Administration of Road Transport
Chang'an Daxue Xuebao (Ziran Kexue Ban)/Journal of Chang'an University (Natural Science Edition) | Year: 2010
In order to evaluate road alignment, the driver direction control model was established based on trajectory-preview theory and real time vehicle state, and the vehicle turning characteristics and operating lag was considered. The computer program was developed with VB 6.0, and simulation experiment was conducted. The result indicated that the model can calculate lateral acceleration and steering wheel angle of test vehicle during driving, the indices reflects safety and driver comfort of the road and can be employed in road alignment evaluation. Remarkable difference of running results between initial speed of 0 and 54 km/h appears only at the start of the road section. During steady running of the experiment, the variation of lateral acceleration and turning angle of steering wheel are in basically accordance with horizontal curve.
Yu H.,Key Laboratory of Urban ITS Technology Optimization and Integration Ministry of Public Security |
Yu H.,Beihang University |
Hu Y.,Beihang University |
Guo H.,Anhui Keli Information Industry Co.
Lecture Notes in Electrical Engineering | Year: 2016
Vehicle queue length is one of the important traffic parameters in intelligent traffic management system. High-altitude video monitoring avoids environmental object barrier and has advantages of dynamic video monitoring like larger view range, multi-angle and high precision, at the same time, it provides technical support for the vehicle queue length detection at the intersection. In order to detect the vehicle queue length in real time and apply it to the management of intelligent traffic system, a new algorithm based on video vehicle queue length detection is proposed in this paper. First, the frame difference method is used to construct the background of the image so that the background modeling error of moving objects could be reduced. On this basis, relief operation is taken to the image background and the current frame image in order to avoid the impact of light changes on the algorithm. Finally, the two-value image is analyzed to obtain the real queue length. The experimental results show that the improved method is simple to achieve and it can obtain a more accurate queue length. © Springer Science+Business Media Singapore 2016.
Gharieh K.,Rutgers University |
Farzan F.,Rutgers University |
Jafari M.A.,Rutgers University |
Gang T.,Anhui Keli Information Industry Co.
21st World Congress on Intelligent Transport Systems, ITSWC 2014: Reinventing Transportation in Our Connected World | Year: 2014
This paper represents traffic conflict technique as a surrogate for real crashes that can be used to develop safety models in intersections. It focuses on pedestrian and crossing/merging vehicles safety in intersections. The high cost of pedestrian crashes to society is the main reason to study pedestrian safety. Training safety models with near miss events would be beneficial due to the fact that frequency of pedestrian crashes is low. In the proposed methodology, a pedestrian's risk is presented as a binomial categorical variable (low risk and high risk). The model is applied to each pedestrian movement at the intersection, so each crosswalk can be compared. By applying the statistical significance test, pedestrian flow and vehicle flow turns out to be the both dominant factors affecting the pedestrians' safety. In addition, finding out the vehicles hot spot locations in an intersection is of great importance. In this paper, a regression model is presented for the number of vehicle-vehicle near misses. Crossing vehicles flow turns out to be more effective on the near miss occurrence compared to the merging vehicles flow. Therefore, the results show the necessity of applying a countermeasure (i.e. protected left turn) to enhance the intersection safety.
Wang Y.,Chang'an University |
Li L.,Shenzhen Urban Transport Planning Center Co. |
Wang Z.,Transportation Institute |
Lv T.,Anhui Ke Li Information Industry Ltd. |
Wang L.,Chang'an University
Journal of Urban Planning and Development | Year: 2013
Xi'an, as a typical inland city in China with a worse integrated transport system, has promoted the use of metro service to ease serious traffic congestion. This paper uses a logistic regression model to examine mode shifts behavior for auto, taxi, bus, electric bicycle, and bicycle users after the implementation of metro service based on stated preference data. The results indicate that auto travelers located in suburban regions are more willing to shift to metro for work trips. Female taxi and auto users are more likely to use metro than males. Longer trips for taxi and electric bicycle travelers prefer to choose the newly introduced metro. Additionally, a preference survey on newly opened metro concluded that metro passengers that shift from auto mode may decrease 8 to 19% because of incomplete transfer facilities. Finally, it was found that the impacts on easing traffic congestion by a single metro corridor are not significant, and some parallel policies need to be adopted. Furthermore, these findings are more useful for developing cities lacking modal joint and integration. © 2013 American Society of Civil Engineers.
Gang T.,Chang'an University |
Gang T.,Key Laboratory Of City Traffic Management Intgn And Optimization Technology Of The Ministy Of Public Security |
Song H.-S.,Chang'an University |
Yan Y.-G.,Anhui Keli Information Industry Co. |
Jafari M.,Transportation Group
Proceedings - 2015 IEEE 12th International Conference on Ubiquitous Intelligence and Computing, 2015 IEEE 12th International Conference on Advanced and Trusted Computing, 2015 IEEE 15th International Conference on Scalable Computing and Communications, 2015 IEEE International Conference on Cloud and Big Data Computing, 2015 IEEE International Conference on Internet of People and Associated Symposia/Workshops, UIC-ATC-ScalCom-CBDCom-IoP 2015 | Year: 2015
The causes of traffic accidents are complex and uncertain, which is difficult to be represented by one or two factors. In order to extract core factors which affect traffic accident and quantify the influences of factors, this paper introduced uncertainty analysis method, rough set theory. Firstly, the information decision table of rough set was formulated based on historical accident data, then the simplified algorithms of rough set model was used to calculate degrees of attributes importance of different factors to their corresponding accident morphologies. Finally, we get the influence degrees of each factor on corresponding traffic accidents morphologies to provide basis of selecting scientific and reasonable indexes for the prediction model of road traffic accident morphologies. © 2015 IEEE.
Hu K.,CAS Hefei Institutes of Physical Science |
Zheng K.,CAS Hefei Institutes of Physical Science |
Tian X.,CAS Hefei Institutes of Physical Science |
Chen L.,CAS Hefei Institutes of Physical Science |
And 2 more authors.
Journal of Macromolecular Science, Part B: Physics | Year: 2011
Attapulgite (AT) was modified by grafting with butyl acrylate (BA) via polymerizations initiated by Gamma radiation. Polypropylene (PP)/AT nanocomposites were synthesized via melt extrusion in a twin-screw extruder. Fourier transform infrared (FTIR) spectroscopy and thermogravimetry (TG) were used to assess the structure of the hybrid materials and the dispersion of AT was verified by transmission electron microscopy (TEM). The crystallization kinetics of PP/AT nanocomposites were investigated by differential scanning calorimetry (DSC) and analyzed by using the Avrami method. The isothermal crystallization kinetics showed that the addition of AT increased both the crystallization rate and the isothermal Avrami exponent of PP. Step-scan differential scanning calorimetry (SDSC) was used to study the influence of AT on the crystallization and subsequent melting behavior. The results revealed that PP and PP/AT nanocomposites experienced multiple melting and secondary crystallization processes during heating. The melting behaviors of PP and PP/AT nanocomposites varied with the variation of crystallization temperature and AT content. Copyright © Taylor & Francis Group, LLC.
Gao W.-B.,Anhui Keli Information Industry Co. |
Zou J.,Anhui Keli Information Industry Co. |
Wu J.,Anhui Keli Information Industry Co.
Jiaotong Yunshu Xitong Gongcheng Yu Xinxi/Journal of Transportation Systems Engineering and Information Technology | Year: 2013
Ramp metering is an effective way to alleviate freeway congestion. It also makes some contributions to increase traffic flow, improve operational efficiency, and reduce traffic accidents, etc. With the rapid growth of the national economy, the public concerns more on the increasing environmental pollution. The study on ramp control considering emissions reduction is an important issue related to people's livelihood. The algorithm in the paper considers some emissions factor. It occupies two objectives, the primary is to minimize total vehicle travel time, and the other is to reduce ramp emission as much as possible. The PARAMICS simulation program is used to evaluate its performance comparing with the ALINEA and no metering condition, and the CMEM traffic emission model is used to calculate the result. Analysis shows that the new ramp metering algorithm is effective in reducing the total vehicle travel time, as well as significantly reducing on-ramp emissions and queues than the ALINEA strategy. Especially, on the large uphill gradient condition, new algorithm could reduce much more total emission. Copyright © 2011 by Science Press.
Gao W.-B.,Anhui Keli Information Industry Co. |
Wu J.,Anhui Keli Information Industry Co. |
Zou J.,Anhui Keli Information Industry Co.
Jiaotong Yunshu Xitong Gongcheng Yu Xinxi/Journal of Transportation Systems Engineering and Information Technology | Year: 2012
With the continual increase of urban traffic demand, traffic congestion occurs more and more frequently. Ramp metering, by regulating the vehicles entering freeway, has been proved to be an effective measure of relieving freeway congestion. This paper presents an optimal ramp metering algorithm with full consideration of the dynamic freeway and ramp density. The algorithm aims to maintain freeway density near the critical value and minimize the ramp queues. The PARAMICS is used to evaluate its performance comparing with the ALINEA and Demand-Capacity algorithms. The data of freeway density, ramp queues, total vehicle travel time and downstream flow are also collected for each simulation. Analysis result shows that the Density-based algorithm is effective in maximizing mainline traffic flow, optimizing the balance of network traffic condition and reducing ramp queues.