English

Using Convolutional Neural Networks to Count Palm Trees in Satellite Images

Computer Vision and Pattern Recognition 2017-01-24 v1

Abstract

In this paper we propose a supervised learning system for counting and localizing palm trees in high-resolution, panchromatic satellite imagery (40cm/pixel to 1.5m/pixel). A convolutional neural network classifier trained on a set of palm and no-palm images is applied across a satellite image scene in a sliding window fashion. The resultant confidence map is smoothed with a uniform filter. A non-maximal suppression is applied onto the smoothed confidence map to obtain peaks. Trained with a small dataset of 500 images of size 40x40 cropped from satellite images, the system manages to achieve a tree count accuracy of over 99%.

Keywords

Cite

@article{arxiv.1701.06462,
  title  = {Using Convolutional Neural Networks to Count Palm Trees in Satellite Images},
  author = {Eu Koon Cheang and Teik Koon Cheang and Yong Haur Tay},
  journal= {arXiv preprint arXiv:1701.06462},
  year   = {2017}
}
R2 v1 2026-06-22T17:57:22.837Z