The Winning Solution to the iFLYTEK Challenge 2021 Cultivated Land Extraction from High-Resolution Remote Sensing Image

TL;DR

This paper presents a highly effective pipeline for accurately extracting cultivated land from high-resolution remote sensing images, achieving first place in the iFLYTEK Challenge 2021.

Contribution

The authors developed a novel tile-based segmentation pipeline with an overlap-tile fusion strategy tailored for remote sensing images, outperforming existing methods.

Findings

Achieved first place in the challenge with top performance.

Developed an effective tile-based segmentation approach.

Proposed a novel overlap-tile fusion strategy.

Abstract

Extracting cultivated land accurately from high-resolution remote images is a basic task for precision agriculture. This report introduces our solution to the iFLYTEK challenge 2021 cultivated land extraction from high-resolution remote sensing image. The challenge requires segmenting cultivated land objects in very high-resolution multispectral remote sensing images. We established a highly effective and efficient pipeline to solve this problem. We first divided the original images into small tiles and separately performed instance segmentation on each tile. We explored several instance segmentation algorithms that work well on natural images and developed a set of effective methods that are applicable to remote sensing images. Then we merged the prediction results of all small tiles into seamless, continuous segmentation results through our proposed overlap-tile fusion strategy. We achieved the first place among 486 teams in the challenge.