Compound geospatial object detection in an aerial image

This paper introduces a knowledge based approach that can be used for the identification of jetty/bridge locations in aerial imagery. With the proposed method, the semantic network formalism to represent declarative knowledge embodied in a jetty/bridge image and the appropriate procedural knowledge, the control procedure was established. A knowledge based system was then introduced through image analysis and interpretation, aiming at accurately locating the desired objects from primary vague identification. With the advanced image processing techniques proposed here, the complexity of using knowledge based system for image analysis is reduced and the proposed method can effectively locate the compound geospatial objects of jetties and bridges.