Vehicle Detection and Classification based on Deep Neural Network for Intelligent Transportation Applications

Tsai, Chia-Chi; Tseng, Cheng-Lung; Tang, Ho-Chia; Guo, Jiun-In

doi:10.23919/apsipa.2018.8659542

Cited by 33 publications

(23 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“… Vehicle lamp detection results with high‐speed vehicles (a) Results of the proposed algorithm, (b) Results obtained by Guo et al [18], (c) Results obtained by Tang and Hussain [24], (d) Results obtained by Tsai et al [30]…”

Section: Experimental Results and Comparisonmentioning

confidence: 99%

“…For a fair and objective analysis, two evaluation methods – pixel‐ and object‐based methods – were adopted in this study. The vehicle lamp detection capability of the proposed algorithm was compared with that of several other algorithms [16, 18, 20, 24, 30].…”

Section: Experimental Results and Comparisonmentioning

confidence: 99%

“…Deep learning technology can automatically extract image features, providing good results in object detection. Therefore, many studies [29–32] have used deep learning for night‐time vehicle detection. Nevertheless, there are only a few data sets and labelling conventions available to train the night‐time vehicle detection algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

High‐accuracy vehicle lamp detection for real‐time night‐time traffic surveillance

Tsai

Chen

2020

IET intell. transp. syst

View full text Add to dashboard Cite

Vehicle lamps are an important image feature of the night‐time vehicle detection algorithm. This study proposes a real‐time night‐time vehicle detection algorithm based on light attenuation characteristic analysis, which consists of vehicle lamp detection and pairing. For the detection phase, this study proposes an automatic dual‐threshold method to quickly extract the attenuation regions around bright objects. This method is highly adaptable and can accurately extract attenuation regions to identify vehicle lamps and strong reflections. For the pairing phase, this study uses the moving trajectories of vehicle lamps to complete preliminary pairing. Then, the vehicle lamp pairing is optimised using the relative vertical and horizontal lamp positions in continuous images. Pixel‐ and object‐based methods were used to verify the experimental vehicle lamp detection results. The results of both verification methods indicate that the proposed algorithm without morphological operation is superior to other algorithms. The accuracy of the pairing phase was >94.6% in scenes with multiple lamps, high‐driving speed, high traffic flow, and rain. Finally, the proposed night‐time vehicle detection algorithm can achieve a real‐time execution speed for images 960 × 720 pixels in size.

show abstract

Section: Experimental Results and Comparisonmentioning

confidence: 99%

Section: Experimental Results and Comparisonmentioning

confidence: 99%

See 1 more Smart Citation

High‐accuracy vehicle lamp detection for real‐time night‐time traffic surveillance

Tsai

Chen

2020

IET intell. transp. syst

View full text Add to dashboard Cite

show abstract

“…After application of specific algorithms, these sensors can be easily used for almost full traffic surveillance, which involves vehicle classification, estimation of vehicle speed or traffic jam detection, eventually as a part of specific sensor systems [ 26 ]. After the implementation of appropriate software, magnetometer sensors could also provide a value-added contribution into the area of traffic surveillance, where other technologies take the role, as for example in sensing drunken drivers [ 27 ], as a part of driving assistance systems [ 28 ], for the edge traffic flow detection [ 29 ], as energy-efficient substitution of cameras for the detection and classification of road vehicles [ 30 ], or for the traffic abnormality detection [ 31 ] etc.…”

Section: Related Workmentioning

confidence: 99%

Which Digital-Output MEMS Magnetometer Meets the Requirements of Modern Road Traffic Survey?

Hodoň

Karpis

Ševčı́k

et al. 2021

Sensors

View full text Add to dashboard Cite

Present systems for road traffic surveillance largely utilize MEMS magnetometers for the purpose of vehicle detection and classification. Magnetoresistive sensing or LR oscillation circuitry are technologies providing the sensors with the competitive advantage which lies in the energy efficiency and low price. There are several chip suppliers on the market who specialize in the development of these sensors. The aim of this paper is to compare available sensors from the viewpoint of their suitability for traffic measurements. A summary of the achieved results is given in the form of the score for each sensor. The introduced sensor chart should provide the audience with knowledge about pros and cons of sensors, especially if intended for the purposes of road traffic surveillance. The authors in this research focused on the specific situation of road traffic monitoring with magnetometers placed at the roadside.

show abstract

“…Deep learning CNN and image-collecting tools, like surveillance cameras, have been combined and used for vehicle classification or real-time traffic monitoring [37][38][39][40]. CNN model improvements, like adopting the layer skipping strategy for better vehicle classification [41], or using CNN-TL to achieve both detection and classification of vehicles that include dump trucks, cars, and buses [42][43][44], have been performed. Apart from vehicle classification or detection, CNN is also able to recognize the working or idle state of earthwork machines, like excavators or trucks [45], and CNN-TL can benefit earthmoving operations or related construction management [27,46].…”

Section: Vison-based Deep Learning In Related Areasmentioning

confidence: 99%

Deep Learning Model Comparison for Vision-Based Classification of Full/Empty-Load Trucks in Earthmoving Operations

et al. 2019

View full text Add to dashboard Cite

Earthmoving is an integral civil engineering operation of significance, and tracking its productivity requires the statistics of loads moved by dump trucks. Since current truck loads’ statistics methods are laborious, costly, and limited in application, this paper presents the framework of a novel, automated, non-contact field earthmoving quantity statistics (FEQS) for projects with large earthmoving demands that use uniform and uncovered trucks. The proposed FEQS framework utilizes field surveillance systems and adopts vision-based deep learning for full/empty-load truck classification as the core work. Since convolutional neural network (CNN) and its transfer learning (TL) forms are popular vision-based deep learning models and numerous in type, a comparison study is conducted to test the framework’s core work feasibility and evaluate the performance of different deep learning models in implementation. The comparison study involved 12 CNN or CNN-TL models in full/empty-load truck classification, and the results revealed that while several provided satisfactory performance, the VGG16-FineTune provided the optimal performance. This proved the core work feasibility of the proposed FEQS framework. Further discussion provides model choice suggestions that CNN-TL models are more feasible than CNN prototypes, and models that adopt different TL methods have advantages in either working accuracy or speed for different tasks.

show abstract

Vehicle Detection and Classification based on Deep Neural Network for Intelligent Transportation Applications

Cited by 33 publications

References 6 publications

High‐accuracy vehicle lamp detection for real‐time night‐time traffic surveillance

High‐accuracy vehicle lamp detection for real‐time night‐time traffic surveillance

Which Digital-Output MEMS Magnetometer Meets the Requirements of Modern Road Traffic Survey?

Deep Learning Model Comparison for Vision-Based Classification of Full/Empty-Load Trucks in Earthmoving Operations

Contact Info

Product

Resources

About