Product Details
Detekce ploch a vodorovného dopravního značení s využitím metod hlubokého učení
Created: 2023
English title
Detection of surfaces and horizontal road markings using deep learning methods
Type
software
License
required - licence fee
Authors
Musil Petr, Ing., Ph.D. (COGNITECHNA)
Klíma Ondřej, Ing., Ph.D. (COGNITECHNA)
Kula Michal, Ing., Ph.D. (DCGM FIT BUT)
Klíma Ondřej, Ing., Ph.D. (COGNITECHNA)
Kula Michal, Ing., Ph.D. (DCGM FIT BUT)
Keywords
deep learning, horizontal road marking, vectorization
Description
Aggregation of data from LiDAR sensors, GPS, IMU and omnidirectional RGB cameras within the Ladybug and Pegasus Leica systems, detection of surfaces (operational road surface, sidewalk, parking area,...) and horizontal traffic markings using deep learning methods, acceleration of time-critical parts using OpenCL/NVidia CUDA. Vectorization of detected surfaces and lines, export of georeferenced results to a map.