Semantic segmentation is in-demand in satellite imagery processing. Because of the complex environment, automatic categorization and segmentation of land cover is a challenging problem. Solving it can help to overcome many obstacles in urban planning, environmental engineering or natural landscape monitoring. In this paper, we propose an approach for automatic multi-class land segmentation based on a fully convolutional neural network of feature pyramid network (FPN) family. This network is consisted of pre-trained on ImageNet Resnet50 encoder and neatly developed decoder. Based on validation results, leaderboard score and our own experience this network shows reliable results for the DEEPGLOBE - CVPR 2018 land cover classification sub-challenge. Moreover, this network moderately uses memory that allows using GTX 1080 or 1080 TI video cards to perform whole training and makes pretty fast predictions.
@article{arxiv.1806.03510,
title = {Feature Pyramid Network for Multi-Class Land Segmentation},
author = {Selim S. Seferbekov and Vladimir I. Iglovikov and Alexander V. Buslaev and Alexey A. Shvets},
journal= {arXiv preprint arXiv:1806.03510},
year = {2018}
}