Automatic detection of transmission towers

The detection of electrical transmission towers in digital images is an essential component of future, automated power networks inspection solutions. In this paper, a novel detection method is proposed. It is composed of two phases: training and detection. In the training phase, the characteristic model of transmission towers is learned from a training set. The training set consists of images with known locations for all transmission towers. In the detection phase, meaningful image features are compared with the model in order to detect likely locations of transmission towers. Experimental results with images downloaded from the Internet show accurate detection even in the presence of background clutter and multiple objects.