Deep Metadata Fusion for Traffic Light to Lane Assignment

Langenberg, Tristan; Lüddecke, Timo; Wörgötter, Florentin

doi:10.1109/lra.2019.2893446

Cited by 6 publications

(7 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The convolutional neural network (CNN) can be utilized not only for the image data sets but also employs metadata in the form of heterogeneous deep data sets [ 1 ] and could assign all of the traffic light information related to the lane information system. To accomplish this, the general CNN topology is trained by converting the input images to predict down-sampling data sets and index vectors [ 2 ].…”

Section: The Related Workmentioning

confidence: 99%

Intelligent Recognition Technology for the Segmentation of Traffic Indication Images Concerning Different Pavement Materials

Xie¹,

Shi²,

Wang³

2022

Applied Bionics and Biomechanics

View full text Add to dashboard Cite

Traffic indication is an important part of the road environment, providing information about road conditions, restrictions, prohibitions, warnings, and the current status related to the flow of the traffic and other navigational aspects. The shape, color, and pictogram of a traffic indication are encoded into the visual characteristics of traffic signs. Not paying attention to these traffic signs could lead directly or indirectly to traffic accidents. In this article, the support traffic indication vector recognition (STIVR) method is proposed to classify the best signal detection to avoid traffic congestion and accidents. The proposed STIVR recognizes the traffic indication system automatically, reduces occurrences of traffic accidents, and helps drivers move safely on different pavement materials. Besides, the adaptive median filter (AMF) algorithm is used to pre-process and protect the traffic indication images without obscuring them. Thus, it indicates the edge of the non-smoothed nasty ferment from the service. In the detection of traffic events, indication images are enhanced, pre-treated, and divided according to symbols and their characteristics such as color, shape, or both. The output becomes a segmented image, including the available space identified as a road sign. The experimental results show that the proposed method functions well; achieves a sufficiently higher process speed and better segmentation of traffic indications and more accuracy in recognition of the objects. For example, the proposed method reaches a higher sensitivity performance of 96%.

show abstract

Section: The Related Workmentioning

confidence: 99%

Intelligent Recognition Technology for the Segmentation of Traffic Indication Images Concerning Different Pavement Materials

Xie¹,

Shi²,

Wang³

2022

Applied Bionics and Biomechanics

View full text Add to dashboard Cite

show abstract

“…Although a vehicle navigation system provides drivers with options for alternative routes based on historical and real-time traffic information, this does not ease traffic congestion owing to the vast number of vehicles navigating during rush hours. Researchers have therefore also adopted adaptive traffic light control systems to coordinate the traffic flow at arterial road intersections [11][12]14]. Nevertheless, there is still much room for the improvement of traffic light control strategies because only local traffic scheduling is typically considered, with the result that the larger scale traffic congestion problem, particularly for a whole city, is not resolved effectively.…”

Section: Introductionmentioning

confidence: 99%

Congestion-Preventing Routing and Charging Scheduling Mechanism for Electric Vehicles in Dense Urban Areas

Huang

et al. 2021

ITC

View full text Add to dashboard Cite

Traffic congestion in metropolitan areas all over the world has become a critical issue that governments mustdeal with effectively. Traffic congestion during rush hours causes vehicle drivers to arrive late at their destinations,resulting in significant economic losses. Although researchers have proposed solutions to the traffic congestionproblem, little research work has presented a joint route and charging planning strategy for electric vehicles(EVs) that alleviates traffic congestion problems simultaneously. Accordingly, a congestion-preventing route and charging planning mechanism for EVs is proposed in this work to tackle the complicated route and charging optimizationproblems of EVs. The route and charging planning proposed in this work analyzes the information providedby EVs, the charging points, and road traffic information simultaneously, and mediates the traffic jammingby means of a route and charging reservation mechanism. Possible occurrence of traffic congestion is detectedin advance and traffic regulation is carried out by allocating an elastic range to the traveling period for late-bookingEVs, to avoid moving during rush hours. EV owners are also encouraged to provide rideshare services forlate-booking EV users during rush hours. The simulation results reveal that the proposed work can satisfy thepreferred route and charging demands of EV users and alleviate traffic congestion effectively.

show abstract

“…The following contents of this chapter 4 were taken identically from the original publication, see [7], and were adapted to the format of my dissertation.…”

Section: Brief Discussion Of the Ieee Its Papermentioning

confidence: 99%

“…Moreover, the deep metadata fusion approach could be extended to a full end-to- 38 Link to the video of the IEEE RA-L paper [7]: https://ieeexplore.ieee.org/ielx7/7083369/ 8581687/8613841/ieee_ra_l_video.mp4?tp=&arnumber=8613841.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning Metadata Fusion for Traffic Light to Lane Assignment

Langenberg¹

View full text Add to dashboard Cite

erzielen die gemessene menschliche Leistung und der tiefe Metadatenfusionsansatz eine nahezu identische Richtig-Positive Rate von 92,9 ±1,3 %. Außerdem wird festgestellt, dass eine frühe Fusion der Metadaten am effektivsten ist und alle fusionierten Metadatenmerkmalkarten sich positiv auf die Ergebnisse auswirken. Ferner hat sich gezeigt, dass der beste Fusionsoperator die elementweise Multiplikation ist. In Anlehnung an die menschliche Wahrnehmung ist festzustellen, dass sich die Ergebnisse mit sinkendem Abstand zur Haltelinie steigern.

show abstract

Deep Metadata Fusion for Traffic Light to Lane Assignment

Cited by 6 publications

References 31 publications

Intelligent Recognition Technology for the Segmentation of Traffic Indication Images Concerning Different Pavement Materials

Intelligent Recognition Technology for the Segmentation of Traffic Indication Images Concerning Different Pavement Materials

Congestion-Preventing Routing and Charging Scheduling Mechanism for Electric Vehicles in Dense Urban Areas

Deep Learning Metadata Fusion for Traffic Light to Lane Assignment

Contact Info

Product

Resources

About