基于改进PSPnet的无人机农田场景语义分割

刘尚旺; 张杨杨; 蔡同波; .

引用本文:	刘尚旺,,张杨杨,,蔡同波, 等..基于改进PSPnet的无人机农田场景语义分割[J].灌溉排水学报,2022,41(4):101-108.
	LIU Shangwang,ZHANG Yangyang,CAI Tongbo, et al..基于改进PSPnet的无人机农田场景语义分割[J].灌溉排水学报,2022,41(4):101-108.

【打印本页】【下载PDF全文】【查看/发表评论】【EndNote】【RefMan】【BibTex】

←前一篇|后一篇→

过刊浏览高级检索

本文已被：浏览 1005次下载 3206次	码上扫一扫！
分享到：微信更多字体:加大+\|默认\|缩小-
基于改进PSPnet的无人机农田场景语义分割
刘尚旺, 张杨杨, 蔡同波, 等.
1.河南师范大学计算机与信息工程学院，河南新乡 453007； 2.中国农业科学院农田灌溉研究所，河南新乡 453002； 3.“智慧商务与物联网技术”河南省工程实验室，河南新乡 453007

摘要:

【目的】改进PSPnet语义分割模型在无人机农田场景下的性能。【方法】对PSPnet语义分割模型进行3方面改进：①通过不同维度特征级联，在强化场景解析的基础上保留更多图像细节特征。②利用深度可分离卷积模块构建轻量级语义分割模型，使其更加高效。③改进激活函数，提升模型分割效果。【结果】所建模型的平均像素准确率和平均交并比分别为89.48%和82.38%，比改进前的模型提高了18.12%和18.93%，且分割结果优于Unet和DeeplabV3+等模型。【结论】改进后的模型能够有效进行无人机遥感农田场景语义分割。

关键词: PSPnet；语义分割；特征级联；深度可分离卷积；激活函数

DOI：10.13522/j.cnki.ggps.2021406

分类号:

基金项目:

An Improved PSPnet Model for Semantic Segmentation of UAV Farmland Images

LIU Shangwang, ZHANG Yangyang, CAI Tongbo, et al.

1. College of computer and information engineering, Henan Normal University, Xinxiang 453007, China; 2. Farmland Irrigation Research Institute, Chinese Academy of Agricultural Sciences, Xinxiang 453002, China; 3. Henan Engineering Laboratory of ‘Smart Business and Internet of Things Technology’, Xinxiang 453007, China

Abstract:

【Background and objective】Deep Learning is an important method in artificial intelligence but has limited applications in agriculture. This paper aims to fill this technology gap based on farmland images acquired using unmanned aerial vehicle (UAV) and the PSPnet segmentation method. 【Method】Different dimensional features in the UAV images were concatenated based on the principle of preserving as many detailed image features as possible via the enhanced scene analysis. A lightweight semantic segmentation model was built using the deep separable convolution module. Finally, the activation function was replaced to improve the segmentation effect of the model. 【Result】Experimental results show that the mean pixel accuracy and the mean intersection over the union of our proposed method are 89.48% and 82.36%, respectively, and their associated segmentation accuracy was improved by 18.12% and 18.93%, respectively. Overall, the segmentation of the proposed method was better than that of Unet and DeeplabV3+.【Conclusion】The proposed method can effectively segment the farmland images acquired by UAV.

Key words: PSPnet; semantic segmentation; feature concatenate; deep separable convolution; activation function