89747-Images-Vehicle-Attributes-Annotation-Data

Description

89,747 Images Vehicle Attributes Annotation Data. The data includes outdoor roads (highway, crossroad) scenes. The data covers multiple vehicle types, multiple vehicle colors, multiple license plate colors, multiple vehicle brands, different time, different vehicle orientations. For a vehicle, 2 rectangular bounding boxes and 5 labels were annotated, the rectangular bounding boxes include the front end or rear end of vehicle, and the whole vehicle; the labels included vehicle color, vehicle type, vehicle brand, vehicle orientation and shooting time; For a vehicle plate, a rectangular bounding box of license plate, the number of license plate and the color of license plate were annotated. This data can be used for tasks such as vehicle attributes analysis and license plate recognition.

For more details, please refer to the link: https://www.nexdata.ai/datasets/121?source=Github

Specifications

Data size

89,747 images, including 26,504 images of speeding scenes and 63,243 images of retrograde scenes

Collecting environment

outdoor roads (highway, crossroad)

Data diversity

including multiple vehicle types, multiple vehicle colors, multiple license plate colors, multiple vehicle brands, different time，different vehicle orientations

Device

surveillance camera, looking down angle

Collecting time

day, night

Image parameter

the image format is .jpg, the annotated file format is .json

Annotation content

for a vehicle, 2 rectangular bounding boxes and 5 labels were annotated, the rectangular bounding boxes include the front end or rear end of vehicle, and the whole vehicle; the labels included vehicle color, vehicle type, vehicle brand,vehicle orientation and shooting time; for a vehicle plate, a rectangular bounding box of license plate, the number of license plate (the number of license plate was replaced with six ‘*’) and the color of license plate were annotated

Accuracy

a rectangular bounding box is qualified when the deviation is not more than 3 pixels, and the qualified rate of the rectangular bounding box shall not be lower than 96%; the accuracy of labels of license plate number annotation, license plate color, vehicle color, vehicle type, vehicle brand, shooting time and vehicle orientation is not less than 96%

Licensing Information

Commercial License