Mapping Plastic Mulched Farmland for High Resolution Images of Unmanned Aerial Vehicle Using Deep Semantic Segmentation

Qinchen Yang,Shuqin Yang,Wenting Han,Man Liu,Zhitao Zhang,Jifeng Ning

doi:10.3390/rs11172008

Abstract

With increasing consumption, plastic mulch benefits agriculture by promoting crop quality and yield, but the environmental and soil pollution is becoming increasingly serious. Therefore, research on the monitoring of plastic mulched farmland (PMF) has received increasing attention. Plastic mulched farmland in unmanned aerial vehicle (UAV) remote images due to the high resolution, shows a prominent spatial pattern, which brings difficulties to the task of monitoring PMF. In this paper, through a comparison between two deep semantic segmentation methods, SegNet and fully convolutional networks (FCN), and a traditional classification method, Support Vector Machine (SVM), we propose an end-to-end deep-learning method aimed at accurately recognizing PMF for UAV remote sensing images from Hetao Irrigation District, Inner Mongolia, China. After experiments with single-band, three-band and six-band image data, we found that deep semantic segmentation models built via single-band data which only use the texture pattern of PMF can identify it well; for example, SegNet reaching the highest accuracy of 88.68% in a 900 nm band. Furthermore, with three visual bands and six-band data (3 visible bands and 3 near-infrared bands), deep semantic segmentation models combining the texture and spectral features further improve the accuracy of PMF identification, whereas six-band data obtains an optimal performance for FCN and SegNet. In addition, deep semantic segmentation methods, FCN and SegNet, due to their strong feature extraction capability and direct pixel classification, clearly outperform the traditional SVM method in precision and speed. Among three classification methods, SegNet model built on three-band and six-band data obtains the optimal average accuracy of 89.62% and 90.6%, respectively. Therefore, the proposed deep semantic segmentation model, when tested against the traditional classification method, provides a promising path for mapping PMF in UAV remote sensing images.

Highlights

Plastic mulching is a method of covering the farmland surface with agricultural plastic membrane of different thicknesses and colors and at different intervals
We found that the texture pattern of plastic mulched farmland (PMF) for Unmanned aerial vehicle (UAV) remote sensing imaging is the determinant factor of plastic film monitoring
The high resolution of UAV remote sensing images makes a spatial pattern of PMF, clearly distinct from other components

Summary

Introduction

Plastic mulching is a method of covering the farmland surface with agricultural plastic membrane of different thicknesses and colors and at different intervals It can stimulate the growth of seedlings and help low-yield or no-yield wasteland to be used for modern agricultural activities by reducing soil moisture evaporation, preventing diseases and pests and maintaining temperature [1]. A huge amount of residues have caused a reduction in land renewable ability and pollution to water and soil, which in turn affect the production of crops [3] To solve these problems, the relevant government departments and enterprises are urgently looking for solutions with experts pushing measures to raise the amount of plastic film recovery [4]. PMF mapping for UAV remote images has very rarely been reported

Methods

Results

Discussion

Conclusion