Article
Remote Sensing
Xiaozhe Zhou, Minfeng Xing, Binbin He, Jinfei Wang, Yang Song, Jiali Shang, Chunhua Liao, Min Xu, Xiliang Ni
Summary: This paper proposed a ground point fitting method based on UAV SfM point cloud to estimate the height of winter wheat. By designing a canopy slice filter and fitting missing ground points, the extraction of ground points was successfully optimized and terrain undulations were effectively simulated to improve accuracy.
Article
Remote Sensing
Yang Song, Jinfei Wang, Bo Shan
Summary: The study enhances the prediction accuracy of winter wheat yield by modifying the SAFY-height model. It establishes the relationship between crop height and biomass using a piecewise linear regression model and calibrates parameters to improve the accuracy of yield estimation for areas with LAI higher than 1.01 m(2)/m(2).
Article
Environmental Sciences
Wenxia Dai, Qingfeng Guan, Shangshu Cai, Rundong Liu, Ruibo Chen, Qing Liu, Chao Chen, Zhen Dong
Summary: This study investigates the performance of estimating plot canopy cover using Unmanned-Aerial-Vehicle (UAV)-Borne Laser Scanning (ULS) and Terrestrial Laser Scanning (TLS). The results show that ULS has better accuracy compared to TLS, and the difference increases with forest complexity. The study provides useful information for the selection of data sources and estimation methods in plot canopy cover mapping.
Article
Environmental Sciences
Quan Yin, Yuting Zhang, Weilong Li, Jianjun Wang, Weiling Wang, Irshad Ahmad, Guisheng Zhou, Zhongyang Huo
Summary: This study aims to optimize UAV flight strategies, incorporate multiple feature selection methods and machine learning algorithms, and enhance the accuracy of SPAD value estimation for different wheat varieties across different growth stages. The study found that the flight altitude significantly impacts the estimation accuracy, with 40 m achieving better results than 20 m.
Article
Agronomy
Xiaokai Chen, Fenling Li, Botai Shi, Kai Fan, Zhenfa Li, Qingrui Chang
Summary: This study explored the feasibility of estimating canopy chlorophyll content (CCC) in winter wheat using a combination of machine learning and canopy spectral transformation (CST). The results showed that the first derivative spectrum (FDS) and continuum removal spectrum (CRS) had a stronger correlation with CCC compared to the original spectrum (OS). Among the parametric regression methods, the univariate regression with CRS-NDSI as the independent variable achieved satisfactory results in CCC estimation. The random forest (RF) regression combined with multiple independent variables had the best accuracy in estimating winter wheat CCC. Therefore, this modeling method could be used as a basic approach for CCC prediction in the Guanzhong Plain area.
Article
Agriculture, Multidisciplinary
Xuan Zhang, Hui Sun, Xingxing Qiao, Xiaobin Yan, Meichen Feng, Lujie Xiao, Xiaoyan Song, Meijun Zhang, Fahad Shafiq, Wude Yang, Chao Wang
Summary: This study optimized vegetation indices for estimating the canopy chlorophyll content (CCC) of winter wheat using different spectral processing methods. The results showed that the first derivative processing improved the correlation between the indices and crop quality traits, and the three-band index improved the monitoring accuracy of CCC.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Environmental Sciences
Shilei Li, Fangjie Li, Maofang Gao, Zhaoliang Li, Pei Leng, Sibo Duan, Jianqiang Ren
Summary: The study introduced a winter wheat mapping method based on singular value decomposition, which improved mapping accuracy by reconstructing spectra based on crop growth characteristics and achieved high overall accuracy and Kappa values in tests in China. Compared to traditional spectral similarity methods, this approach avoids excessive information extraction and has advantages in distinguishing non-winter wheat pixels.
Article
Environmental Sciences
Mahendra Bhandari, Shannon Baker, Jackie C. Rudd, Amir M. H. Ibrahim, Anjin Chang, Qingwu Xue, Jinha Jung, Juan Landivar, Brent Auvermann
Summary: The study revealed that drought significantly limits wheat productivity. By utilizing unmanned aerial systems to collect data, it is possible to effectively monitor the effects of drought on wheat growth and productivity.
Article
Plant Sciences
Narendra S. Chandel, Yogesh A. Rajwade, Kumkum Dubey, Abhilash K. Chandel, A. Subeesh, Mukesh K. Tiwari
Summary: Timely detection of crop water stress can help improve irrigation management and reduce yield loss. This study utilized computer vision and thermal-RGB imagery to monitor water stress in winter wheat for two years. The results showed that thermal imagery outperformed RGB imagery in classification accuracy, indicating its potential in high-throughput mitigation and management of crop water stress.
Article
Agronomy
Yixin Sun, Yusen Luo, Qian Zhang, Lizhang Xu, Liying Wang, Pengpeng Zhang
Summary: This study proposes a method for estimating the distribution of rice plant height based on a moving surface and 3D point cloud elevation. By utilizing the statistical outlier removal algorithm and moving surface fitting elevation, accurate classification of ground and crop point cloud data is achieved, resulting in high accuracy and low error rates.
Article
Chemistry, Analytical
Haitao Li, Gengchen Wu, Shutian Tao, Hao Yin, Kaijie Qi, Shaoling Zhang, Wei Guo, Seishi Ninomiya, Yue Mu
Summary: This paper proposes an automatic branch-leaf segmentation pipeline based on lidar point cloud for nondestructive and accurate measurements of leaf phenotypic parameters. The proposed method establishes a 3D model and uses the PointNet++ model for segmentation, achieving efficient and convenient measurements compared to manual methods.
Article
Environmental Sciences
Fangjie Li, Jianqiang Ren, Shangrong Wu, Hongwei Zhao, Ningdan Zhang
Summary: This study focused on the winter wheat mapping method based on remote sensing technology, proposing a new method to improve the accuracy of crop area extraction, achieving a high consistency with statistical data.
Article
Agronomy
Ling Zheng, Qun Chen, Jianpeng Tao, Yakun Zhang, Yu Lei, Jinling Zhao, Linsheng Huang
Summary: This study evaluated the ability of spectral features, image textures, and their combinations to estimate winter wheat aboveground biomass (AGB). The results showed that the combined use of spectral and image textures can effectively improve the accuracy of AGB estimation at all growth stages.
Article
Agronomy
Marie Therese Abi Saab, Razane El Alam, Ihab Jomaa, Sleiman Skaf, Salim Fahed, Rossella Albrizio, Mladen Todorovic
Summary: The integration of remote sensing technology and crop growth models was effective in monitoring winter wheat growth in Lebanon's Bekaa Valley. By incorporating vegetation indices from Sentinel 2 satellite into the AquaCrop model, biomass and yield of wheat were accurately simulated under different water regimes, showing promising results for crop growth prediction in Mediterranean areas.
Article
Environmental Sciences
Xiangyang Liu, Yaxiong Wang, Feng Kang, Yang Yue, Yongjun Zheng
Summary: This study obtained LiDAR point cloud data for Citrus grandis var. Longanyou and developed a canopy reconstruction method, achieving accurate calculation of canopy volume. The ASBS algorithm performed the best, while the CH algorithm had the shortest computation time, providing theoretical reference for future modules such as accurate plant protection, orchard obstacle avoidance, and biomass estimation.
Article
Chemistry, Analytical
Morten Stigaard Laursen, Rasmus Nyholm Jorgensen, Henrik Skov Midtiby, Kjeld Jensen, Martin Peter Christiansen, Thomas Mosgaard Giselsson, Anders Krogh Mortensen, Peter Kryger Jensen
Article
Chemistry, Analytical
Esma Mujkic, Mark P. Philipsen, Thomas B. Moeslund, Martin P. Christiansen, Ole Ravn
Summary: This study poses the object detection problem in autonomous agricultural vehicles as anomaly detection and applies convolutional autoencoders to identify objects that deviate from the normal pattern. The results show that the semisupervised autoencoder (SSAE) outperforms other autoencoder models in detecting unknown objects and is comparable to the YOLOv5-based object detector. Additionally, SSAE is capable of detecting unknown objects, whereas the object detector fails to do so.
Article
Robotics
Esma Mujkic, Ole Ravn, Martin Peter Christiansen
Summary: This paper presents an ensemble method that combines semantic segmentation, object detection, and anomaly detection tasks in agricultural scenes. The proposed method detects agriculture-specific classes using an object detector and detects other objects using an anomaly detector. The segmentation map is utilized to provide additional information about the location of objects. The results show that combining object detection with anomaly detection increases the number of detected objects in agricultural scene images.
FRONTIERS IN ROBOTICS AND AI
(2023)
Article
Chemistry, Analytical
Esma Mujkic, Martin P. Christiansen, Ole Ravn
Summary: Vision-based object detection is crucial for autonomous agricultural vehicles, but the limited availability of labeled datasets in the agricultural domain poses a challenge. This paper tackles this challenge by proposing two YOLOv5-based object detection models, one pre-trained on a large-scale dataset for general object detection and another trained on a smaller set of agriculture-specific classes. To improve inference, the authors propose an ensemble module based on a hierarchical structure of classes. Experimental results show that the proposed module increases mAP@.5 from 0.575 to 0.65 on the test dataset and reduces misclassification of similar classes detected by different models. Furthermore, by translating detections to a higher level in the class hierarchy, the overall mAP@.5 can be increased to 0.701 at the cost of reduced class granularity.
Proceedings Paper
Engineering, Electrical & Electronic
Tobias Lundby, Martin Peter Christiansen, Kjeld Jensen
2019 INTERNATIONAL CONFERENCE ON UNMANNED AIRCRAFT SYSTEMS (ICUAS' 19)
(2019)
Article
Computer Science, Information Systems
Martin Peter Christiansen, Nima Teimouri, Morten Stigaard Laursen, Birgitte Feld Mikkelsen, Rasmus Nyholm Jorgensen, Claus Aage Gron Sorensen
Article
Robotics
Martin Peter Christiansen, Peter Gorm Larsen, Rasmus Nyholm Jorgensen
Article
Robotics
Gareth Edwards, Martin P. Christiansen, Dionysis D. Bochtis, Claus G. Sorensen