Object Detection Based on Feature Extraction and Morphological Operations

An object detection method for aerial images based on feature extraction and mathematical morphology is proposed in this paper. The changes of the background have smaller influences on this method with less computational complexity. The method is composed of three parts: initial segmentation, morphology processing and object detection. Initial segmentation can find rough object regions by feature extracting and fuzzy c-means clustering. The following morphology processing identifies the correct locations and boundaries of all objects in the original image. The last stage of detection is achieved by means of the automatic threshold method. The experimental result shows that it is valid and especially suitable for detecting objects from the background full of lawn