DoubleNet: A Method for Generating Navigation Lines of Unstructured Soil Roads in a Vineyard Based on CNN and Transformer

Navigating unstructured roads in vineyards with weak satellite signals presents significant challenges for robotic systems. This research introduces DoubleNet, an innovative deep-learning model designed to generate navigation lines for such conditions. To improve the model’s ability to extract image...

Full description

Saved in:
Bibliographic Details
Main Authors: Xuezhi Cui, Licheng Zhu, Bo Zhao, Ruixue Wang, Zhenhao Han, Kunlei Lu, Xuguang Feng, Jipeng Ni, Xiaoyi Cui
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Agronomy
Subjects:
Online Access:https://www.mdpi.com/2073-4395/15/3/544
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850203950468300800
author Xuezhi Cui
Licheng Zhu
Bo Zhao
Ruixue Wang
Zhenhao Han
Kunlei Lu
Xuguang Feng
Jipeng Ni
Xiaoyi Cui
author_facet Xuezhi Cui
Licheng Zhu
Bo Zhao
Ruixue Wang
Zhenhao Han
Kunlei Lu
Xuguang Feng
Jipeng Ni
Xiaoyi Cui
author_sort Xuezhi Cui
collection DOAJ
description Navigating unstructured roads in vineyards with weak satellite signals presents significant challenges for robotic systems. This research introduces DoubleNet, an innovative deep-learning model designed to generate navigation lines for such conditions. To improve the model’s ability to extract image features, DoubleNet incorporates several key innovations, such as a unique multi-head self-attention mechanism (Fused-MHSA), a modified activation function (SA-GELU), and a specialized operation block (DNBLK). Based on them, DoubleNet is structured as an encoder–decoder network that includes two parallel subnetworks: one dedicated to processing 2D feature maps and the other focused on 1D tensors. These subnetworks interact through two feature fusion networks, which operate in both the encoder and decoder stages, facilitating a more integrated feature extraction process. Additionally, we utilized a specially annotated dataset comprising images fused with RGB and mask, with five navigation points marked to enhance the accuracy of point localization. As a result of these innovations, DoubleNet achieves a remarkable 95.75% percentage of correct key points (PCK) and operates at 71.16 FPS on our dataset, with a combined performance that outperformed several well-known key point detection algorithms. DoubleNet demonstrates strong potential as a competitive solution for generating effective navigation routes for robots operating in vineyards with unstructured roads.
format Article
id doaj-art-fbde68a2f9244d13bfc05e7007d1e2b0
institution OA Journals
issn 2073-4395
language English
publishDate 2025-02-01
publisher MDPI AG
record_format Article
series Agronomy
spelling doaj-art-fbde68a2f9244d13bfc05e7007d1e2b02025-08-20T02:11:24ZengMDPI AGAgronomy2073-43952025-02-0115354410.3390/agronomy15030544DoubleNet: A Method for Generating Navigation Lines of Unstructured Soil Roads in a Vineyard Based on CNN and TransformerXuezhi Cui0Licheng Zhu1Bo Zhao2Ruixue Wang3Zhenhao Han4Kunlei Lu5Xuguang Feng6Jipeng Ni7Xiaoyi Cui8State Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaState Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaState Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaState Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaState Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaState Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaState Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaState Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaState Key Laboratory of Agricultural Equipment Technology, Beijing 100083, ChinaNavigating unstructured roads in vineyards with weak satellite signals presents significant challenges for robotic systems. This research introduces DoubleNet, an innovative deep-learning model designed to generate navigation lines for such conditions. To improve the model’s ability to extract image features, DoubleNet incorporates several key innovations, such as a unique multi-head self-attention mechanism (Fused-MHSA), a modified activation function (SA-GELU), and a specialized operation block (DNBLK). Based on them, DoubleNet is structured as an encoder–decoder network that includes two parallel subnetworks: one dedicated to processing 2D feature maps and the other focused on 1D tensors. These subnetworks interact through two feature fusion networks, which operate in both the encoder and decoder stages, facilitating a more integrated feature extraction process. Additionally, we utilized a specially annotated dataset comprising images fused with RGB and mask, with five navigation points marked to enhance the accuracy of point localization. As a result of these innovations, DoubleNet achieves a remarkable 95.75% percentage of correct key points (PCK) and operates at 71.16 FPS on our dataset, with a combined performance that outperformed several well-known key point detection algorithms. DoubleNet demonstrates strong potential as a competitive solution for generating effective navigation routes for robots operating in vineyards with unstructured roads.https://www.mdpi.com/2073-4395/15/3/544orchard navigationunstructured roadconvolutionFused-MHSA
spellingShingle Xuezhi Cui
Licheng Zhu
Bo Zhao
Ruixue Wang
Zhenhao Han
Kunlei Lu
Xuguang Feng
Jipeng Ni
Xiaoyi Cui
DoubleNet: A Method for Generating Navigation Lines of Unstructured Soil Roads in a Vineyard Based on CNN and Transformer
Agronomy
orchard navigation
unstructured road
convolution
Fused-MHSA
title DoubleNet: A Method for Generating Navigation Lines of Unstructured Soil Roads in a Vineyard Based on CNN and Transformer
title_full DoubleNet: A Method for Generating Navigation Lines of Unstructured Soil Roads in a Vineyard Based on CNN and Transformer
title_fullStr DoubleNet: A Method for Generating Navigation Lines of Unstructured Soil Roads in a Vineyard Based on CNN and Transformer
title_full_unstemmed DoubleNet: A Method for Generating Navigation Lines of Unstructured Soil Roads in a Vineyard Based on CNN and Transformer
title_short DoubleNet: A Method for Generating Navigation Lines of Unstructured Soil Roads in a Vineyard Based on CNN and Transformer
title_sort doublenet a method for generating navigation lines of unstructured soil roads in a vineyard based on cnn and transformer
topic orchard navigation
unstructured road
convolution
Fused-MHSA
url https://www.mdpi.com/2073-4395/15/3/544
work_keys_str_mv AT xuezhicui doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer
AT lichengzhu doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer
AT bozhao doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer
AT ruixuewang doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer
AT zhenhaohan doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer
AT kunleilu doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer
AT xuguangfeng doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer
AT jipengni doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer
AT xiaoyicui doublenetamethodforgeneratingnavigationlinesofunstructuredsoilroadsinavineyardbasedoncnnandtransformer