Real-time on-road vehicle and motorcycle detection using a single camera

Duan, Bobo; Liu, Wei; Fu, Pengyu; Yang, Chunyang; Wen, Xuezhi; Yuan, Huai

doi:10.1109/icit.2009.4939585

Cited by 32 publications

(26 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The approaches above may not be able to detect motorcycles and bicycles. Therefore, in [97], [98], along with edges, shadow and black Fig. 11.…”

Section: ) Motion Based Approachmentioning

confidence: 99%

“…Therefore, it is easier to identify and track a motorcycle among the vehicles. However, the algorithms proposed in [97], [98] might fail to detect motorcycles during nighttime since the shadow information used is no longer applicable.…”

Section: F Analysis On Motorcycle Detectionmentioning

confidence: 99%

“…The classifier was integrated with a tracking system (seen in session IV) where motorcyclists were automatically segmented from video data. Techniques proposed in [97], [98] showed successful classification and tracking of motorcycles on urban roads. Prior information was used to find out the possible candidates for potential motorcycles.…”

Section: B Verification Stagementioning

confidence: 99%

“…Majority of existing on-road vehicle detection techniques [97], [98] use detect-then-track approach (i.e., vehicles are detected before being tracked). Although, it is possible to build a CAS without tracking by running and relying on detection schemes for each frame but integration of tracking with detection techniques is productive due to following reasons.…”

Section: Analysis On Trackingmentioning

confidence: 99%

See 3 more Smart Citations

Vehicle Detection Techniques for Collision Avoidance Systems: A Review

Mukhtar

Xia

Tang

2015

IEEE Trans. Intell. Transport. Syst.

375

202

View full text Add to dashboard Cite

Over the past decade, vision-based vehicle detection techniques for road safety improvement have gained an increasing amount of attention. Unfortunately, the techniques suffer from robustness due to huge variability in vehicle shape (particularly for motorcycles), cluttered environment, various illumination conditions, and driving behavior. In this paper, we provide a comprehensive survey in a systematic approach about the state-of-the-art on-road vision-based vehicle detection and tracking systems for collision avoidance systems (CASs). This paper is structured based on a vehicle detection processes starting from sensor selection to vehicle detection and tracking. Techniques in each process/step are reviewed and analyzed individually. Two main contributions in this paper are the following: survey on motorcycle detection techniques and the sensor comparison in terms of cost and range parameters. Finally, the survey provides an optimal choice with a low cost and reliable CAS design in vehicle industries.

show abstract

“…The approaches above may not be able to detect motorcycles and bicycles. Therefore, in [97], [98], along with edges, shadow and black Fig. 11.…”

Section: ) Motion Based Approachmentioning

confidence: 99%

Section: F Analysis On Motorcycle Detectionmentioning

confidence: 99%

Section: B Verification Stagementioning

confidence: 99%

Section: Analysis On Trackingmentioning

confidence: 99%

See 2 more Smart Citations

Vehicle Detection Techniques for Collision Avoidance Systems: A Review

Mukhtar

Xia

Tang

2015

IEEE Trans. Intell. Transport. Syst.

375

202

View full text Add to dashboard Cite

show abstract

“…Therefore, motorcycle detection is becoming very important, specially in developing countries, where they are involved in a large share of the traffic and so high accident rates are reported, due to the high vulnerability of this kind of vehicle. Until now most motorcycle video detection systems are implemented using conventional feature extraction techniques such as 3D models [1] [2], vehicle dimensions [3] [4] [5], symmetry, color, shadow, geometrical features and texture and even wheel contours [6]. Other works report the use of stable features [7], HOG for evaluation of helmet presence [8] [9], variations of HOG [10] [11], and the use of SIFT, DSIFT and SURF [12].…”

Section: Introductionmentioning

confidence: 99%

Motorcycle Classification in Urban Scenarios using Convolutional Neural Networks for Feature Extraction

Espinosa¹,

Velastín²,

Branch³

2017

8th International Conference of Pattern Recognition Systems (ICPRS 2017)

View full text Add to dashboard Cite

This paper presents a motorcycle classification system for urban scenarios using Convolutional Neural Network (CNN). Significant results on image classification has been achieved using CNNs at the expense of a high computational cost for training with thousands or even millions of examples. Nevertheless, features can be extracted from CNNs already trained. In this work AlexNet, included in the framework CaffeNet, is used to extract features from frames taken on a real urban scenario. The extracted features from the CNN are used to train a support vector machine (SVM) classifier to discriminate motorcycles from other road users. The obtained results show a mean accuracy of 99.40% and 99.29% on a classification task of three and five classes respectively. Further experiments are performed on a validation set of images showing a satisfactory classification.

show abstract

Detection and Tracking of Motorcycles in Congested Urban Environments Using Deep Learning and Markov Decision Processes

Espinosa

Velastín

Branch

2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

This research describes "EspiNet", a Deep Learning Convolutional Neural Network model, in conjunction with a Markov Decision Process (MDP) tracker for detection and tracking of occluded motorcycles in urban environments. The model is trained and evaluated, using a new public dataset with up to 10,000 annotated images, created for this research, and captured in real urban traffic scenes. Images were captured using a moving camera mounted in a drone, where more than 60% of the motorcycles are affected by occlusions. The network design involves many tests, where a promising result of 88.84% in average precision (AP) is achieved, despite the considerable number of occluded vehicles, the movement of the camera and the low angle used for capture. The model predictions are used as input to an MDP tracker, reaching results up to 85.2% in Multiple Object Tracking Accuracy (MOTA). The proposed network architecture outperforms state of the art YOLO (You Look Only Once) v3.0 and Faster R-CNN (VGG16 based) detection models, producing also better tracking results in comparison with the use of the other two models as detector base for the MDP tracker.

show abstract

Real-time on-road vehicle and motorcycle detection using a single camera

Cited by 32 publications

References 15 publications

Vehicle Detection Techniques for Collision Avoidance Systems: A Review

Vehicle Detection Techniques for Collision Avoidance Systems: A Review

Motorcycle Classification in Urban Scenarios using Convolutional Neural Networks for Feature Extraction

Detection and Tracking of Motorcycles in Congested Urban Environments Using Deep Learning and Markov Decision Processes

Contact Info

Product

Resources

About