Abstract

TORRIJOS C., Germán. Classification of types of paviments for urban roads from an Ortoimagen and object-orient analysis. Rev. Téc. Ing. Univ. Zulia [online]. 2013, vol.36, n.3, pp.230-239. ISSN 0254-0770.

The vehicular traffic in Bogotá City, capital of Colombia, is very high due to convergences of all transport system and constitutes by itself the commercial, cultural and industrial heartland of the country. With city’s urban growth and economic development of the city, the number of vehicles circulating inside it is increasing year after year, mainly in the metropolitan region. Due to this increase, its been observed a deterioration in the city’s highway system, making necessary to search alternatives that can mitigate this problem. In this context, the main purpose of this research is to classify the types of Bogota’s pavements of urban roads, using orthoimages produced by the Geographic Institute “Agustín Codazzi” -IGAC, in order to find an alternative to map the roads that need to be restored. To evaluate the method, a study was developed in small downtown area using the OBIA (Object Based Image Analysis) implemented in Ecognition software. In the final results two technologies of evaluation were used: the counterfoil of confusion across the calculation of statistical coefficients and the analysis of measures of uncertainty from the statistics of stability of the classification, giving like proved the production of a thematic map with the types of pavements of the zone of study with a precision result of 58,19%, determined by the coefficient of agreement Kappa (k), which is considered to be a good qualification.

Keywords : analysis of image-based objects; object-oriented classification; OBIA.