Recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusion

IntroductionRatoon rice, as a high-efficiency rice cultivation mode, is widely applied around the world. Mechanical righting of rolled rice stubble can significantly improve yield in regeneration season, but lack of automation has become an important factor restricting its further promotion.MethodsI...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuanrui Li, Liping Xiao, Zhaopeng Liu, Muhua Liu, Peng Fang, Xiongfei Chen, Jiajia Yu, Jinlong Lin, Jinping Cai
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-01-01
Series:Frontiers in Plant Science
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fpls.2025.1533206/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832576360658239488
author Yuanrui Li
Yuanrui Li
Liping Xiao
Liping Xiao
Zhaopeng Liu
Zhaopeng Liu
Muhua Liu
Muhua Liu
Peng Fang
Peng Fang
Xiongfei Chen
Xiongfei Chen
Jiajia Yu
Jiajia Yu
Jinlong Lin
Jinlong Lin
Jinping Cai
Jinping Cai
author_facet Yuanrui Li
Yuanrui Li
Liping Xiao
Liping Xiao
Zhaopeng Liu
Zhaopeng Liu
Muhua Liu
Muhua Liu
Peng Fang
Peng Fang
Xiongfei Chen
Xiongfei Chen
Jiajia Yu
Jiajia Yu
Jinlong Lin
Jinlong Lin
Jinping Cai
Jinping Cai
author_sort Yuanrui Li
collection DOAJ
description IntroductionRatoon rice, as a high-efficiency rice cultivation mode, is widely applied around the world. Mechanical righting of rolled rice stubble can significantly improve yield in regeneration season, but lack of automation has become an important factor restricting its further promotion.MethodsIn order to realize automatic navigation of the righting machine, a method of fusing an instance segmentation model and a monocular depth prediction model was used to realize monocular localization of the rolled rice stubble rows in this study.ResultsTo achieve monocular depth prediction, a depth estimation model was trained on training set we made, and absolute relative error of trained model on validation set was only 7.2%. To address the problem of degradation of model's performance when migrated to other monocular cameras, based on the law of the input image’s influence on model's output results, two optimization methods of adjusting inputs and outputs were used that decreased the absolute relative error from 91.9% to 8.8%. After that, we carried out model fusion experiments, which showed that CD (chamfer distance) between predicted 3D coordinates of navigation points obtained by fusing the results of the two models and labels was only 0.0990. The CD between predicted point cloud of rolled rice stubble rows and label was only 0.0174.
format Article
id doaj-art-b4751941ca2347a1993aea15a8204cac
institution Kabale University
issn 1664-462X
language English
publishDate 2025-01-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Plant Science
spelling doaj-art-b4751941ca2347a1993aea15a8204cac2025-01-31T06:41:04ZengFrontiers Media S.A.Frontiers in Plant Science1664-462X2025-01-011610.3389/fpls.2025.15332061533206Recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusionYuanrui Li0Yuanrui Li1Liping Xiao2Liping Xiao3Zhaopeng Liu4Zhaopeng Liu5Muhua Liu6Muhua Liu7Peng Fang8Peng Fang9Xiongfei Chen10Xiongfei Chen11Jiajia Yu12Jiajia Yu13Jinlong Lin14Jinlong Lin15Jinping Cai16Jinping Cai17College of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaCollege of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaCollege of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaCollege of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaCollege of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaCollege of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaCollege of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaCollege of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaCollege of Engineering, Jiangxi Agricultural University, Nanchang, ChinaJiangxi Key Laboratory of Modern Agricultural Equipment, Nanchang, ChinaIntroductionRatoon rice, as a high-efficiency rice cultivation mode, is widely applied around the world. Mechanical righting of rolled rice stubble can significantly improve yield in regeneration season, but lack of automation has become an important factor restricting its further promotion.MethodsIn order to realize automatic navigation of the righting machine, a method of fusing an instance segmentation model and a monocular depth prediction model was used to realize monocular localization of the rolled rice stubble rows in this study.ResultsTo achieve monocular depth prediction, a depth estimation model was trained on training set we made, and absolute relative error of trained model on validation set was only 7.2%. To address the problem of degradation of model's performance when migrated to other monocular cameras, based on the law of the input image’s influence on model's output results, two optimization methods of adjusting inputs and outputs were used that decreased the absolute relative error from 91.9% to 8.8%. After that, we carried out model fusion experiments, which showed that CD (chamfer distance) between predicted 3D coordinates of navigation points obtained by fusing the results of the two models and labels was only 0.0990. The CD between predicted point cloud of rolled rice stubble rows and label was only 0.0174.https://www.frontiersin.org/articles/10.3389/fpls.2025.1533206/fullratoon ricemodel fusiondepth predictiondeep learningmonocular vision
spellingShingle Yuanrui Li
Yuanrui Li
Liping Xiao
Liping Xiao
Zhaopeng Liu
Zhaopeng Liu
Muhua Liu
Muhua Liu
Peng Fang
Peng Fang
Xiongfei Chen
Xiongfei Chen
Jiajia Yu
Jiajia Yu
Jinlong Lin
Jinlong Lin
Jinping Cai
Jinping Cai
Recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusion
Frontiers in Plant Science
ratoon rice
model fusion
depth prediction
deep learning
monocular vision
title Recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusion
title_full Recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusion
title_fullStr Recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusion
title_full_unstemmed Recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusion
title_short Recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusion
title_sort recognition and localization of ratoon rice rolled stubble rows based on monocular vision and model fusion
topic ratoon rice
model fusion
depth prediction
deep learning
monocular vision
url https://www.frontiersin.org/articles/10.3389/fpls.2025.1533206/full
work_keys_str_mv AT yuanruili recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT yuanruili recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT lipingxiao recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT lipingxiao recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT zhaopengliu recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT zhaopengliu recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT muhualiu recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT muhualiu recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT pengfang recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT pengfang recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT xiongfeichen recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT xiongfeichen recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT jiajiayu recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT jiajiayu recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT jinlonglin recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT jinlonglin recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT jinpingcai recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion
AT jinpingcai recognitionandlocalizationofratoonricerolledstubblerowsbasedonmonocularvisionandmodelfusion