Automatic building identification and features extraction from aerial images: Application on the historic 1866 square of Chania Greece

During the last decade significant efforts have been made to monitor cities and record their current situation and growth through time. Advances in aerial photography and the implementation of ortho-images enabled the cartography of large scale areas leading to numerous applications in data recording in both rural and urban regions. Features extraction from such images poses a challenging goal especially in areas exhibiting little chromatic variations, such as multiple detached cement buildings surrounded by cement pavements and tar roads, and dynamic noise conditions due to large building shades and moving vehicles. This paper presents a fully automated software solution for the identification of compact buildings-blocks from aerial ortho-photographs. The aim is to identify and record compact building-blocks´ perimeters ignoring any lesser features displayed by them in order to produce a set of planar coordinates for every building-block. This information is most valuable for topographical city mapping applications and can be further exploited towards city-modeling either by stereo-photography or laser scanning technology, and necessary scaling and alignment techniques to import the height information in an expanded spatial set of coordinates.