Volume XLII-1
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-1, 19-23, 2018
https://doi.org/10.5194/isprs-archives-XLII-1-19-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-1, 19-23, 2018
https://doi.org/10.5194/isprs-archives-XLII-1-19-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

  26 Sep 2018

26 Sep 2018

SEGMENT-AND-COUNT: VEHICLE COUNTING IN AERIAL IMAGERY USING ATROUS CONVOLUTIONAL NEURAL NETWORKS

S. Azimi, E. Vig, F. Kurz, and P. Reinartz S. Azimi et al.
  • German Aerospace Center (DLR), Remote Sensing Technology Institute, D-82234 Weßling, Germany

Keywords: Vehicle Segmentation, Vehicle Counting, Aerial Imagery, Convolutional Neural Networks, Atrous Convolution

Abstract. High-resolution aerial imagery can provide detailed and in some cases even real-time information about traffic related objects. Vehicle localization and counting using aerial imagery play an important role in a broad range of applications. Recently, convolutional neural networks (CNNs) with atrous convolution layers have shown better performance for semantic segmentation compared to conventional convolutional aproaches. In this work, we propose a joint vehicle segmentation and counting method based on atrous convolutional layers. This method uses a multi-task loss function to simultaneously reduce pixel-wise segmentation and vehicle counting errors. In addition, the rectangular shapes of vehicle segmentations are refined using morphological operations. In order to evaluate the proposed methodology, we apply it to the public “DLR 3K” benchmark dataset which contains aerial images with a ground sampling distance of 13 cm. Results show that our proposed method reaches 81.58 % mean intersection over union in vehicle segmentation and shows an accuracy of 91.12 % in vehicle counting, outperforming the baselines.