Building surface damage recognition based on synthetic data

Abstract

To detect surface damage to buildings, it is necessary to involve workers who are at risk of industrial injuries when inspecting hard-to-reach areas of industrialpremises. Attraction of special means, such as aerial platforms, safety systems, etc. increase the financial costs with this approach. The use of unmanned aerial vehicles, coupled withneural network algorithms, can simplify this procedure. Due tothe inaccessibility, the problem of obtaining training data forneural networks arises, which can be solved by synthesizingthem in a virtual environment.