Exploiting street-level panoramic images for large-scale automated surveying of traffic signs

Hazelhoff, Lykele; Creusen, Ivo; With, Peter H. N. de

doi:10.1007/s00138-014-0628-z

Cited by 15 publications

(26 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With respect to processing time, the traffic sign recognition system takes about 88.5 s per image on average [3]. The road marking recognition system requires 10 s per image on average.…”

Section: Individual Recognition Resultsmentioning

confidence: 99%

“…The road-sign recognition system processes all panoramic images captured within a region of interest, and is described extensively in [3]. Figure 3 displays the system overview.…”

Section: Road-sign Recognition Systemmentioning

confidence: 99%

“…Figure 3 displays the system overview. Each main system component is briefly summarized below, and more extensively in [3].…”

Section: Road-sign Recognition Systemmentioning

confidence: 99%

“…For example, several systems are described for creating surveys of road signs from streetlevel images [1,2,3,4]. Most of such systems focus on the recognition of signs in single images, while only few also identify their 3D positions [2,3].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Combined generation of road marking and road sign databases applied to consistency checking of pedestrian crossings

Hazelhoff¹,

Creusen²,

Woudsma³

et al. 2015

2015 14th IAPR International Conference on Machine Vision Applications (MVA)

Self Cite

View full text Add to dashboard Cite

Combined road marking and traffic sign databases are beneficial for both road maintenance and for usage within navigation devices and autonomous driving vehicles. The combination of both markings and signs completely provides all instructions and legislation for drivers. This paper presents a conceptual system for the automated creation of such combined databases and investigates the benefit of this combination for the specific case of pedestrian crossings. Evaluations on 62 km of road have shown that individual detection of road signs and markings indicating pedestrian crossings is very accurate (≥ 95%), enabling selective safety analysis towards these specific locations. Combining both approaches enables very accurate identification of crosswalks and additionally leads to the retrieval of crossings with undetectable markings or signs, such that maintenance can be directed specifically towards these potentially dangerous crosswalks.

show abstract

“…With respect to processing time, the traffic sign recognition system takes about 88.5 s per image on average [3]. The road marking recognition system requires 10 s per image on average.…”

Section: Individual Recognition Resultsmentioning

confidence: 99%

“…The road-sign recognition system processes all panoramic images captured within a region of interest, and is described extensively in [3]. Figure 3 displays the system overview.…”

Section: Road-sign Recognition Systemmentioning

confidence: 99%

“…Figure 3 displays the system overview. Each main system component is briefly summarized below, and more extensively in [3].…”

Section: Road-sign Recognition Systemmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Combined generation of road marking and road sign databases applied to consistency checking of pedestrian crossings

Hazelhoff¹,

Creusen²,

Woudsma³

et al. 2015

2015 14th IAPR International Conference on Machine Vision Applications (MVA)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Street furniture identification has many applications including road maintenance and urban plannings. Moreover, the government needs position and type information of street furniture in order to maintain it when it's broken or its visibility degrades over time (Hazelhoff et al, 2014). Urban planners require the existing street furniture information to locate the new ones.…”

Section: Introductionmentioning

confidence: 99%

Fully Convolutional Networks for Street Furniture Identification in Panorama Images

Wang

Zhou

et al. 2019

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Panoramic images are widely used in many scenes, especially in virtual reality and street view capture. However, they are new for street furniture identification which is usually based on mobile laser scanning point cloud data or conventional 2D images. This study proposes to perform semantic segmentation on panoramic images and transformed images to separate light poles and traffic signs from background implemented by pre-trained Fully Convolutional Networks (FCN). FCN is the most important model for deep learning applied on semantic segmentation for its end to end training process and pixel-wise prediction. In this study, we use FCN-8s model that pre-trained on cityscape dataset and finetune it by our own data. The results show that in both pre-trained model and fine-tuning, transformed images have better prediction results than panoramic images.

show abstract

Large-Scale Street Space Quality Evaluation Based on Deep Learning Over Street View Image

Liu

Han²,

Xiong

et al. 2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Exploiting street-level panoramic images for large-scale automated surveying of traffic signs

Cited by 15 publications

References 21 publications

Combined generation of road marking and road sign databases applied to consistency checking of pedestrian crossings

Combined generation of road marking and road sign databases applied to consistency checking of pedestrian crossings

Fully Convolutional Networks for Street Furniture Identification in Panorama Images

Large-Scale Street Space Quality Evaluation Based on Deep Learning Over Street View Image

Contact Info

Product

Resources

About