Open AccessMethod for constructing a network of urban roads using satellite images
Author(s) -
E. M. Vikhtenko,
A A Tusikova
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1047/1/012037
Subject(s) - convolutional neural network , transformation (genetics) , computer science , hough transform , satellite , mathematical morphology , artificial intelligence , artificial neural network , computer vision , urban network , pattern recognition (psychology) , remote sensing , data mining , image (mathematics) , image processing , geography , engineering , biochemistry , chemistry , gene , aerospace engineering , environmental planning
An automatic method for recognizing the urban network of highways on high-resolution satellite images has been developed and implemented, based on the use of the Mask-RCNN convolutional neural network. The results of the neural network operation are post-processed by the Hough transformation and algorithms of mathematical morphology. The proposed method has acceptable accuracy and high speed of detection of asphalt urban roads.