Detection of Transmission Line Corridors Risk Intrusion based on BCNN

Risk intrusion of transmission lines is one of the main environmental challenges causing power interruption. Many techniques have been used to detect vegetation erosion. These methods are very effective in detecting vegetation invasion. However, they are expensive as cover areas are concerned. Alternatively, simple surveillance device can overlay vast areas at a relatively low cost. In this paper, we describe the statistical moments of the color space and the texture features of images to identify the most effective features that can improve the accuracy of vegetation density classification of the B-CNN algorithm. The method aims to distinguish between high density and low density vegetation areas along the power line corridor right of way (ROW). The results show that the statistical moments of the color space have a positive effect on the classification accuracy, while certain features of the grayscale symbiotic matrix have a negative effect on the classification accuracy. Therefore, a combination of the most effective features is used to achieve a recall of 98.272%.