Industrial Automatic Control Systems and Controllers Annotation << Back Automatic Detection and Object Classification of Transport Flow Using Data Remote Sensing and of Geographical Information about Roads V.S. Tormozov The article proposes an algorithm scheme for automatic detection and classification of TP objects using space images of ultra-high spatial resolution and geographical OSM of road data on the ground. Problem definition is formulated. An analysis of existing studies on the issue is provided. The scheme is built on the basis of interpolating model of MAC overhangs and neural network method of deep machine learning for recognition and classification of objects. The neural network method was trained on space survey data. The performance of both the entire circuit and each of its constituent algorithms on the actual DDS and OSM data have been evaluated. Keywords: satellite imagery; ultra-high resolution; OSM data; street road network; artificial neural network; digital image processing; machine learning; artificial intelligence; image recognition. DOI: 10.25791/asu.6.2020.1189 Pp. 18-26.