A new method for detecting rooftops in satellite images is presented. The proposed method is based on a combination of machine\r\nlearning techniques, namely, k-means clustering and support vector machines (SVM). Firstly k-means clustering is used to segment\r\nthe image into a set of rooftop candidatesââ?¬â?these are homogeneous regions in the image which are potentially associated with\r\nrooftop areas. Next, the candidates are submitted to a classification stage which determines which amongst them correspond to\r\nââ?¬Å?trueââ?¬Â rooftops. To achieve improved accuracy, a novel two-pass classification process is used. In the first pass, a trained SVM is\r\nused in the normal way to distinguish between rooftop and nonrooftop regions.However, this can be a challenging task, resulting in\r\na relatively high rate of misclassification. Hence, the second pass, which we call the ââ?¬Å?histogram method,ââ?¬Â was devised with the aim\r\nof detecting rooftops which were missed in the first pass. The performance of the model is assessed both in terms of the percentage\r\nof correctly classified candidates as well as the accuracy of the estimated rooftop area.
Loading....