
"CfSM, Car-free Street Mapping" is a cutting-edge product that leverages deep learning and drone technology to automatically eliminate vehicle images from road surfaces.
In July of last year (2022), it was granted the status of a venture startup innovation procurement item by the Public Procurement Service's Venture Nation.
There are two primary rationales for the removal of vehicle images from the road surface. Firstly, by effectively eliminating vehicle images, we can obtain accurate road condition information that may be obscured by these vehicles. This includes crucial data regarding road hazards such as potholes and cracks, which are notorious contributors to traffic accidents, allowing for precise assessment of their size, severity, and precise location. Secondly, by efficiently removing vehicle images, we can accurately extract road lane information, which is often concealed by passing vehicles. Consequently, we are empowered to create highly accurate and detailed high-definition maps (HDMaps) tailored specifically for autonomous vehicles.
The most significant differentiating factor is that with the conventional methods, if there are problems or difficulties with the obtained data, it often requires multiple re-shoots or revisits to resolve them. However, by using our solution, there is no need for re-shooting or revisiting. With just one capture, all these issues can be effectively addressed.