A Fast shape detection approach by directional integrations

Okman, Osman Erman
Detection and identification of objects from aerial images are important problems for various types of application areas. For many of the man-made structures shape is a fundamental feature by which these objects are separated from the background and other structures. In this thesis, a novel geometric shape detection algorithm based on the spatial properties of structures is proposed. Since the objects are transformed into 1-D vectors by evaluating directional integrals and detections occur by the analysis of the features obtained in those 1-D spaces, the proposed approach requires less memory and computation than most of the approaches reported in the literature. Detailed derivation of the method is given and the experimental results are presented in order to show the detection performance of the method under different amount of noise and geometric deformations. Experimental results on real images also show that the proposed approach can significantly speed up the computation without degrading the performance. Moreover, a Petroleum Oil Lubricants (POL) depots identification procedure in high-resolution satellite images is developed where detection of the circular structures is one of the crucial steps, which is achieved by the proposed shape detection approach. Performed experiments over a large data set imply promising identification performance and the usability of the shape detection approach in real world applications.
Citation Formats
O. E. Okman, “A Fast shape detection approach by directional integrations,” Ph.D. - Doctoral Program, Middle East Technical University, 2013.