Headland Identification and Ranging Method for Autonomous Agricultural Machines
Headland boundary identification and ranging are the key supporting technologies for the automatic driving of intelligent agricultural machinery, and they are also the basis for controlling operational behaviors such as autonomous turning and machine lifting. The complex, unstructured environments o...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2024-02-01
|
Series: | Agriculture |
Subjects: | |
Online Access: | https://www.mdpi.com/2077-0472/14/2/243 |
_version_ | 1797299145063530496 |
---|---|
author | Hui Liu Kun Li Luyao Ma Zhijun Meng |
author_facet | Hui Liu Kun Li Luyao Ma Zhijun Meng |
author_sort | Hui Liu |
collection | DOAJ |
description | Headland boundary identification and ranging are the key supporting technologies for the automatic driving of intelligent agricultural machinery, and they are also the basis for controlling operational behaviors such as autonomous turning and machine lifting. The complex, unstructured environments of farmland headlands render traditional image feature extraction methods less accurate and adaptable. This study utilizes deep learning and binocular vision technologies to develop a headland boundary identification and ranging system built upon the existing automatic guided tractor test platform. A headland image annotation dataset was constructed, and the MobileNetV3 network, notable for its compact model structure, was employed to achieve binary classification recognition of farmland and headland images. An improved MV3-DeeplabV3+ image segmentation network model, leveraging an attention mechanism, was constructed, achieving a high mean intersection over union (<i>MIoU)</i> value of 92.08% and enabling fast and accurate detection of headland boundaries. Following the detection of headland boundaries, binocular stereo vision technology was employed to measure the boundary distances. Field experiment results indicate that the system’s average relative errors of distance in ranging at distances of 25 m, 20 m, and 15 m are 6.72%, 4.80%, and 4.35%, respectively. This system is capable of meeting the real-time detection requirements for headland boundaries. |
first_indexed | 2024-03-07T22:46:26Z |
format | Article |
id | doaj.art-3dcbce584ad141c29027ad02504134d4 |
institution | Directory Open Access Journal |
issn | 2077-0472 |
language | English |
last_indexed | 2024-03-07T22:46:26Z |
publishDate | 2024-02-01 |
publisher | MDPI AG |
record_format | Article |
series | Agriculture |
spelling | doaj.art-3dcbce584ad141c29027ad02504134d42024-02-23T15:03:42ZengMDPI AGAgriculture2077-04722024-02-0114224310.3390/agriculture14020243Headland Identification and Ranging Method for Autonomous Agricultural MachinesHui Liu0Kun Li1Luyao Ma2Zhijun Meng3Information Engineering College, Capital Normal University, Beijing 100048, ChinaInformation Engineering College, Capital Normal University, Beijing 100048, ChinaInformation Engineering College, Capital Normal University, Beijing 100048, ChinaNational Research Center of Intelligent Equipment for Agriculture, Beijing 100097, ChinaHeadland boundary identification and ranging are the key supporting technologies for the automatic driving of intelligent agricultural machinery, and they are also the basis for controlling operational behaviors such as autonomous turning and machine lifting. The complex, unstructured environments of farmland headlands render traditional image feature extraction methods less accurate and adaptable. This study utilizes deep learning and binocular vision technologies to develop a headland boundary identification and ranging system built upon the existing automatic guided tractor test platform. A headland image annotation dataset was constructed, and the MobileNetV3 network, notable for its compact model structure, was employed to achieve binary classification recognition of farmland and headland images. An improved MV3-DeeplabV3+ image segmentation network model, leveraging an attention mechanism, was constructed, achieving a high mean intersection over union (<i>MIoU)</i> value of 92.08% and enabling fast and accurate detection of headland boundaries. Following the detection of headland boundaries, binocular stereo vision technology was employed to measure the boundary distances. Field experiment results indicate that the system’s average relative errors of distance in ranging at distances of 25 m, 20 m, and 15 m are 6.72%, 4.80%, and 4.35%, respectively. This system is capable of meeting the real-time detection requirements for headland boundaries.https://www.mdpi.com/2077-0472/14/2/243autonomous agricultural machineryheadlandimage recognitiondeep learningbinocular vision |
spellingShingle | Hui Liu Kun Li Luyao Ma Zhijun Meng Headland Identification and Ranging Method for Autonomous Agricultural Machines Agriculture autonomous agricultural machinery headland image recognition deep learning binocular vision |
title | Headland Identification and Ranging Method for Autonomous Agricultural Machines |
title_full | Headland Identification and Ranging Method for Autonomous Agricultural Machines |
title_fullStr | Headland Identification and Ranging Method for Autonomous Agricultural Machines |
title_full_unstemmed | Headland Identification and Ranging Method for Autonomous Agricultural Machines |
title_short | Headland Identification and Ranging Method for Autonomous Agricultural Machines |
title_sort | headland identification and ranging method for autonomous agricultural machines |
topic | autonomous agricultural machinery headland image recognition deep learning binocular vision |
url | https://www.mdpi.com/2077-0472/14/2/243 |
work_keys_str_mv | AT huiliu headlandidentificationandrangingmethodforautonomousagriculturalmachines AT kunli headlandidentificationandrangingmethodforautonomousagriculturalmachines AT luyaoma headlandidentificationandrangingmethodforautonomousagriculturalmachines AT zhijunmeng headlandidentificationandrangingmethodforautonomousagriculturalmachines |