Abstract
State-of-the-art object detection approaches such as Fast/Faster R-CNN, SSD, or YOLO have difficulties detecting dense, small targets with arbitrary orientation in large aerial images. The main reason is that using interpolation to align RoI features can result in a lack of accuracy or even loss of location information. We present the Local-aware Region Convolutional Neural Network (LR-CNN), a novel two-stage approach for vehicle detection in aerial imagery. We enhance translation invariance to detect dense vehicles and address the boundary quantization issue amongst dense vehicles by aggregating the high-precision RoIs’ features. Moreover, we resample high-level semantic pooled features, making them regain location information from the features of a shallower convolutional block. This strengthens the local feature invariance for the resampled features and enables detecting vehicles in an arbitrary orientation. The local feature invariance enhances the learning ability of the focal loss function, and the focal loss further helps to focus on the hard examples. Taken together, our method better addresses the challenges of aerial imagery. We evaluate our approach on several challenging datasets (VEDAI, DOTA), demonstrating a significant improvement over state-of-the-art methods. We demonstrate the good generalization ability of our approach on the DLR 3K dataset.
Original language | English |
---|---|
Title of host publication | XXIV ISPRS Congress, Commission II 2020 |
Editors | N. Paparoditis, C. Mallet, F. Lafarge, F. Remondino, I. Toschi, T. Fuse |
Place of Publication | Nice |
Publisher | International Society for Photogrammetry and Remote Sensing (ISPRS) |
Pages | 381-388 |
Number of pages | 8 |
DOIs | |
Publication status | Published - 3 Aug 2020 |
Event | XXIVth ISPRS Congress 2020 - Virtual Event, Nice, France Duration: 4 Jul 2020 → 10 Jul 2020 Conference number: 24 http://www.isprs2020-nice.com |
Publication series
Name | ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
---|---|
Publisher | Copernicus |
Volume | V-2-2020 |
ISSN (Print) | 2194-9042 |
Conference
Conference | XXIVth ISPRS Congress 2020 |
---|---|
Abbreviated title | ISPRS 2020 |
Country/Territory | France |
City | Nice |
Period | 4/07/20 → 10/07/20 |
Internet address |