SAIT Traffic Landmark Dataset

Citation Author(s):
Won Hee
Lee
Samsung Advanced Institute of Technology
Kyungboo
Jung
Samsung Advanced Institute of Technology
Chulwoo
Kang
Samsung Advanced Institute of Technology
Hyun Sung
Chang
Samsung Advanced Institute of Technology
Submitted by:
Wonhee Lee
Last updated:
Sat, 01/16/2021 - 21:37
DOI:
10.21227/gar1-z597
Data Format:
License:
64 Views
Categories:
Keywords:
0
0 ratings - Please login to submit your rating.

Abstract 

We disclose a traffic landmark dataset for detection.The dataset generated with our framework includes about 150,000 images and annotations of about 470,000 traffic landmarks.Our dataset was collected in an urban area of Seoul and suburban areas of Suwon, Hwaseong, Yongin, and Seongnam in South Korea at different times of the day.Images taken in the morning or evening included a large number of saturated areas due to exposure to direct sunlight.Most images taken under the light condition of the late evening was low-contrast.The images taken at noon included the reflection of the windshield due to strong sunlight.These various elements contribute to the development of a robust detection algorithm because they can be encountered in a real test environment. 

Instructions: 

Annotated classes are six landmark types:warning sign, prohibition sign, mandatory sign, supplementary sign, traffic light, and road marking.  The location of a traffic landmark is annotated with reference point instead of bounding box. Traffic lights and traffic signs have their center points as reference points.The reference points of road markings are located at the lower right corner.

 

CSV files

  • filename(string): name of image file

 

  • class(integer): class of each traffic landmark 

(0:warning sign, 1:prohibition sign, 2:mandatory sign, 3:supplementary sign, 4;traffic light, 5: road marking)

  • refpt_x, refpt_y(float): image coordinate of reference point 

 

Citation: 

Won Hee Lee, Kyungboo Jung, Chulwoo Kang, and Hyun Sung Chang, "Semi-automatic framework for traffic landmark annotation", IEEE Open Journal of Intelligent Transportation System, 2021.

   

Comments

.

Submitted by Wonhee Lee on Sat, 01/16/2021 - 21:38