Identifying field and road modes of agricultural Machinery based on GNSS Recordings: A graph convolutional neural network approach

Chen, Ying; Li, Guangyuan; Zhang, Xiaoqiang; Jia, Jiepeng; Zhou, Kun; Wu, Caicong

doi:10.1016/j.compag.2022.107082

Cited by 20 publications

(25 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although several field-road classification methods based solely on global navigation satellite system (GNSS) recordings of agricultural machinery have been proposed, they often suffer from either a performance robustness issue or a time consumption problem. The term “performance robustness” refers to the stability of classification performance across various input trajectories, while “time computation” denotes the time required to run a classification method ( Chen et al, 2021 ; Poteko, Eder & Noack, 2021 ; Chen et al, 2022 ; Zhang et al, 2022 ). Given that practical applications require both effective and efficient field-road classification, it is necessary to address these two issues.…”

Section: Introductionmentioning

confidence: 99%

“…In contrast, there is a significant drop in performance on lower-quality trajectories. To address the performance robustness problem, a deep learning method known as GCN was introduced ( Chen et al, 2022 ). The method constructs a spatio-temporal graph based on spatial and temporal relationships between points in an input trajectory and then employs the graph convolution process to extract spatio-temporal features for supporting field-road classification.…”

Section: Introductionmentioning

confidence: 99%

“…These movement-based features have been proven to be helpful for classifying transportation modes. However, these features are not sufficient for the quality classification of agricultural machinery trajectories, as both movement-based and point-density-based characteristics play important roles ( Chen et al, 2021 ; Chen et al, 2022 ; Zhang et al, 2022 ). So, when dealing with the trajectory quality classification task, a dedicated feature extraction method is essential.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Trajectory classification to support effective and efficient field-road classification

Chen,

Kuang,

2024

PeerJ Computer Science

Self Cite

View full text Add to dashboard Cite

Field-road classification, which automatically identifies in-field activities and out-of-field activities in global navigation satellite system (GNSS) recordings, is an important step for the performance evaluation of agricultural machinery. Although several field-road classification methods based only on GNSS recordings have been proposed, there is a trade-off between time consumption and accuracy performance for such methods. To obtain an optimal balance, it is important to choose a suitable field-road classification method for each trajectory based on its GNSS trajectory quality. In this article, a trajectory classification task was proposed, which classifies the quality of GNSS trajectories into three categories (high-quality, medium-quality, or low-quality). Then, a trajectory classification (TC) model was developed to automatically assign a quality category to each input trajectory, utilizing global and local features specific to agricultural machinery. Finally, a novel field-road classification method is proposed, wherein the selection of field-road classification methods depends on the trajectory quality category predicted by the TC model. The comprehensive experiments show that the proposed trajectory classification method achieved 86.84% accuracy, which consistently outperformed current trajectory classification methods by about 2.6%, and the proposed field-road classification method has obtained a balance between efficiency and effectiveness, i.e., sufficient efficiency with a tolerable accuracy loss. This is the first attempt to examine the balance problem between efficiency and effectiveness in existing field-road classification methods and to propose a trajectory classification specific to these methods.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Trajectory classification to support effective and efficient field-road classification

Chen,

Kuang,

2024

PeerJ Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…Nevertheless, the model conducts inadequate and empirical-based feature extraction for trajectory data in the absence of theoretical support. Besides, the generalization performance of a single classifier applied in the model is limited, which is often inferior to that of a multi-classifier system based on integrated learning [18,19] . Chen et al [20] proposed a supervised deep learning model based on graph convolutional neural network (GCN), which constructs a spatio-temporal graph based on temporal and spatial features of trajectory points, and then applies graph convolution to find new feature representations for the trajectory points.…”

Section: Introductionmentioning

confidence: 99%

DR-XGBoost: An XGBoost model for field-road segmentation based on dual feature extraction and recursive feature elimination

Xiao,

Mo,

Xiong

et al. 2023

International Journal of Agricultural and Biological Engineering

View full text Add to dashboard Cite

Field-road segmentation is one of the key tasks in the processing of the trajectory of agricultural machinery. To improve the accuracy of the field-road segmentation, this study proposed an XGBoost model based on dual feature extraction and recursive feature elimination called DR-XGBoost. DR-XGBoost takes only a small amount of agricultural machine trajectory features as input. Firstly, the model adopted the dual feature extraction method we designed to rapidly expand the number of features and then adequately extract local trajectory features by the time window and feature extraction operator. Secondly, the model applies the recursive feature elimination algorithm to eliminate redundant features from the perspective of the model segmentation effect and thus reduce the computational consumption of model training. Thirdly, it trains XGBoost to complete the trajectory segmentation. To evaluate the effectiveness of DR-XGBoost, a series of experiments were conducted on a real trajectory dataset of agricultural machines. The model achieves a 98.2% Macro-F1 score on the dataset, which is 10.9% higher than the previous state-of-art. The proposal of DR-XGBoost fills the knowledge gap of trajectory feature extraction for agricultural machinery and provides a reasonable and effective feature selection scheme for the field-road segmentation problem.

show abstract

“…Though DBSCAN-based algorithms [21,22] can achieve good segmentation accuracy, they need to adjust parameters manually, which means the algorithms are not suitable for big data processing. While for deep learning methods, a graph convolutional network (GCN) algorithm was developed for field & road classification, which achieved 88.14% and 85.93% for the wheat harvesting data and the paddy harvesting data, respectively [23]. The GCN-based algorithm explored a segmentation method with spatial-temporal relationships, but its accuracy was not so high that did not meet the requirement of big data processing.…”

Section: Introductionmentioning

confidence: 99%

Field & Road-CENet: Semantic Segmentation for Agricultural Machinery Trajectory Segmentation

Huang

Wan

2023

Preprint

Self Cite

View full text Add to dashboard Cite

Agricultural machinery management is the key to agricultural production, andtrajectory segmentation lays an important foundation for machinerymanagement. For big data platform of agricultural machinery, it is hoped tosimultaneously improve both of segmentation accuracy and segmentationefficiency to satisfy the processing requirements. However, traditional machinelearning algorithms need to manually adjust parameters and do not integratemultiple features when dealing with trajectory segmentation. This paper aims atthe above problems by modifying the CE-Net model to improve the segmentationaccuracy and ensure the processing efficiency. A creative method for constructingtrajectory image from discrete trajectory data was developed in this paper. Thenew trajectory image showed both the temporal and the spatial characteristics ofthe operation. Then, a semantic segmentation algorithm, Field & Road- CENet,was put forward. The proposed algorithm modified the network CE-Net by addingtwo modules, Standard Convolution Residual Block and Global MaxpoolingAttention Mechanism, to fuse original feature information and enhance thesemantic expression of low-level feature maps. Two cotton sowing datasets werebuilt, including the meter level and the centimeter level. Experiment results showthat Field & Road-CENet performed well on both datasets. In the fieldsegmentation that is the most concerned, the identification accuracy reached97.8% and 95.2%, respectively, and the average accuracy of field and road was94.2% and 88.3%, respectively. In conclusion, this work verifies the feasibility ofusing semantic segmentation to realize trajectory segmentation of agriculturalmachinery. Compared with the current researches, the proposed method isapplicable to trajectory data with two precisions, which has stronger domaingeneralization ability. And it performs quite fast with an average inference time of 0.044 s for each image block, demonstrating that the proposed algorithm issuitable for the big data processing of agricultural machinery.

show abstract

Identifying field and road modes of agricultural Machinery based on GNSS Recordings: A graph convolutional neural network approach

Cited by 20 publications

References 17 publications

Trajectory classification to support effective and efficient field-road classification

Trajectory classification to support effective and efficient field-road classification

DR-XGBoost: An XGBoost model for field-road segmentation based on dual feature extraction and recursive feature elimination

Field & Road-CENet: Semantic Segmentation for Agricultural Machinery Trajectory Segmentation

Contact Info

Product

Resources

About

Identifying field and road modes of agricultural Machinery based on GNSS Recordings: A graph convolutional neural network approach

Cited by 20 publications

References 17 publications

Trajectory classification to support effective and efficient field-road classification

Trajectory classification to support effective and efficient field-road classification

DR-XGBoost: An XGBoost model for field-road segmentation based on dual feature extraction and recursive feature elimination

Field &amp; Road-CENet: Semantic Segmentation for Agricultural Machinery Trajectory Segmentation

Contact Info

Product

Resources

About

Field & Road-CENet: Semantic Segmentation for Agricultural Machinery Trajectory Segmentation