Superpixel-Based Multiscale CNN Approach Toward Multiclass Object Segmentation From UAV-Captured Aerial Images
Tanmay Kumar Behera, Sambit Bakshi, Michele Nappi, Pankaj Kumar
Abstract
Unmanned aerial vehicles (UAVs) are promising remote sensors capable of reforming remote sensing applications. However, for artificial intelligence (AI)-guided tasks, like land cover mapping and ground-object mapping, most deep learning-based architectures fail to extract scale-invariant features, resulting in poor performance accuracy. In this context, the article proposes a superpixel-aided multiscale convolutional neural network (CNN) architecture to avoid misclassification in complex urban aerial images.The proposed framework is a two-tier deep learning-based segmentation architecture. In the first stage, a superpixel-based simple linear iterative cluster (SLIC) algorithm produces superpixel images with crucial contextual information. The second stage comprises a multiscale CNN architecture that uses these information-rich superpixel images to extract scale-invariant features for predicting the object class of each pixel. Two UAV-image-based aerial image datasets: <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">NITRDrone</i> dataset and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">urban drone dataset</i> (UDD) are considered to perform the experimentation. The proposed model outperforms the considered state-of-the-art methods with an intersection of union (IoU) of 76.39% and 86.85% on UDD and NITRDrone datasets, respectively. Experimentally obtained results prove that the proposed architecture performs superior by achieving better performance accuracy in complex and challenging scenarios.