Effect of Spatial Filtering on Object Detection with the SURF Algorithm

Abstract

The article presents a preliminary study into the detection of electrical insulators in digital images aquired during a power line inspection. Due to the enormous amount of digital data generated during a high voltage lines inspection, there is a need to automate the detection process of power insulators in digital images. As part of the study, the effects of applying spatial filtering into digital images for the purpose of the identification of electrical insulators with the use of a local feature detector and descriptor SURF (Speeded Up Robust Features) were analyzed. The recognition and designation of an insulator’s ROI (Region Of Interest) in a digital image will allow the application of more advanced methods aimed at the identification of possible damages in further stages of the analysis.