The Winning Solution to the iFLYTEK Challenge 2021 Cultivated Land Extraction from High-Resolution Remote Sensing Image
This work addresses a domain-specific task in precision agriculture by providing an incremental improvement in cultivated land extraction from remote sensing images.
The authors tackled the problem of extracting cultivated land from high-resolution remote sensing images for precision agriculture, achieving first place among 486 teams in the iFLYTEK Challenge 2021.
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.