The road driving image data consists of archived images taken by our company during regular driving surveys of roads nationwide, using front and rear cameras.
Since 2005, still images have been recorded at approximately 5-meter intervals, preserving a time-series record of the landscape and changes around roads throughout Japan.
The images contain a wealth of information, including road shapes, roadside facilities, intersections, signs, and road structures, making them suitable for a wide range of tasks such as GIS verification, mapping, supplementing field surveys, and year-on-year comparisons.
It can be used in a variety of fields, including infrastructure maintenance and management, navigation system development, AI model training data, and marketing analysis.
By leveraging the accumulation of driving images over many years, we can compare and analyze changes in roads and surrounding environments over time.
It allows for the visualization of changes in signs and structures, as well as the progression of pavement conditions, which can be used for verifying urban development and as evidence.
Images containing a variety of objects along the roadside can be used as training data for object detection AI targeting signs and directional markers. Because location information is included, it is also suitable for visualizing and organizing detection results on a map.
In road infrastructure maintenance, it can be used to check for pavement deterioration and cracks, and to build automated detection models.
This image data can be used in the research and development of autonomous driving technology to improve the accuracy of recognizing intersections and road shapes.
Yes, it covers all prefectures nationwide, and provides comprehensive coverage from main roads to local roads.
Yes, image data that includes a wealth of roadside signs, billboards, directional signs, and structures is used as training data for object detection and image recognition models.
We have a proven track record of implementation, particularly in sign location recognition and road damage detection.
Our company has accumulated driving image data since 2005, and it is possible to extract past images by specifying a location, road section, or prefecture.
If you specify the required time period, we can display and compare past and present images of that location side by side.
The images are provided along with a log of GPS-based latitude and longitude information acquired during driving. This ensures practical spatial accuracy even in urban areas and along roadsides.
Masking (blurring) is applied to identifiable information such as people's faces and vehicle license plates.
We can provide the necessary support.
When providing images, we can offer masked images depending on your intended use.
High-quality, high-precision, and highly up-to-date road data covering approximately 1.3 million km across Japan
Spot data for public facilities nationwide
Includes future information up to three years in advance, such as store openings, relocations, and closures
