Mapping Utility Poles in Aerial Orthoimages Using ATSS Deep Learning Method

Abstract

Mapping utility poles using side-view images acquired with car-mounted cameras is a time-consuming task, mainly in larger areas due to the need for street-by-street surveying. Aerial images cover larger areas and can be feasible alternatives although the detection and mapping of the utility poles in urban environments using top-view images is challenging. Thus, we propose the use of Adaptive Training Sample Selection (ATSS) for detecting utility poles in urban areas since it is a novel method and has not yet investigated in remote sensing applications. Here, we compared ATSS with Faster Region-based Convolutional Neural Networks (Faster R-CNN) and Focal Loss for Dense Object Detection (RetinaNet ), currently used in remote sensing applications, to assess the performance of the proposed methodology. We used 99,473 patches of 256 × 256 pixels with ground sample distance (GSD) of 10 cm. The patches were divided into training, validation and test datasets in approximate proportions of 60%, 20% and 20%, respectively. As the utility pole labels are point coordinates and the object detection methods require a bounding box, we assessed the influence of the bounding box size on the ATSS method by varying the dimensions from 30×30 to 70×70 pixels. For the proposal task, our findings show that ATSS is, on average, 5% more accurate than Faster R-CNN and RetinaNet. For a bounding box size of 40×40, we achieved Average Precision with intersection over union of 50% (AP50) of 0.913 for ATSS, 0.875 for Faster R-CNN and 0.874 for RetinaNet. Regarding the influence of the bounding box size on ATSS, our results indicate that the AP50 is about 6.5% higher for 60×60 compared to 30×30. For AP75, this margin reaches 23.1% in favor of the 60×60 bounding box size. In terms of computational costs, all the methods tested remain at the same level, with an average processing time around of 0.048 s per patch. Our findings show that ATSS outperforms other methodologies and is suitable for developing operation tools that can automatically detect and map utility poles.

Active Fire Mapping on Brazilian Pantanal Based on Deep Learning and CBERS 04A Imagery

“…To summarize, our results indicate that VFNET provided the highest F1-Score, followed by RetinaNet, SABL, Faster R-CNN, and ATSS. Previous studies in remote sensing [49,50] showed that ATSS provided more accurate results for pole and apple detection; however, for active fire detection, ATSS provided less accurate results due to a small rate of True Positives, indicating the inability of the trained model (considering the same number of training epochs of the remaining algorithms) to identify active fire regions.…”

Reidentification of Objects From Aerial Photos With Hybrid Siamese Neural Networks

“…Concerning electric poles in particular, a lot of attention has gone into ways of mapping them periodically, including manned [1] and unmanned [2] aerial flights, and remote sensing pipelines [3] (we refer to [2] for a more in-depth overview). Concerning deep learning instead, deep neural networks have been used to predict possible failures [19], identifying specific poles from images [20], or finding vegetation or icing on the poles [2]. State-of-the-art methods are generally framed as an object detection problem, where the task is to find the proper bounding box surrounding a pole from an aerial image [20].…”

“…Concerning deep learning instead, deep neural networks have been used to predict possible failures [19], identifying specific poles from images [20], or finding vegetation or icing on the poles [2]. State-of-the-art methods are generally framed as an object detection problem, where the task is to find the proper bounding box surrounding a pole from an aerial image [20]. In this article, we consider an intermediate problem, where we assume that poles have been successfully identified in multiple photos (taken from successive aerial routes, see Section III), but we need to reidentify the same pole from different images to plan potential maintenance activities.…”

Utility‐pole Detection Based on Interwoven Column Generation from Terrestrial Mobile Laser Scanner Data

“…The algorithm down samples the point cloud according to voxel size, filters the downsampled point cloud to remove data points corresponding to ground reflections, divides the point cloud into a plurality of clusters that are parallel to the vehicle’s direction of movement, projects the plurality of clusters to a plane that is parallel to the sides of the vehicle to generate an image, detects a pole‐shaped object in the image and tracks the pole‐shaped object by deep sorting across multiple frames of the plurality of frames (Chen et al, 2020). Clustering point clouds in complex city scenes with attached walls is usually time‐consuming and very challenging (Gomes et al, 2020).…”