Computers and Electronics in Agriculture最新文献

{"title":"Assessing the potential of quantum computing in agriculture","authors":"Torsten Pook ,&nbsp;Jeremie Vandenplas ,&nbsp;Juan Carlos Boschero ,&nbsp;Esteban Aguilera ,&nbsp;Koen Leijnse ,&nbsp;Aneesh Chauhan ,&nbsp;Yamine Bouzembrak ,&nbsp;Rob Knapen ,&nbsp;Michael Aldridge","doi":"10.1016/j.compag.2025.110332","DOIUrl":"10.1016/j.compag.2025.110332","url":null,"abstract":"<div><div>With increasing computational demands in agriculture and life sciences, quantum computing is emerging as a potential alternative to classical computing. Unlike classical computers, which utilize binary bits, quantum computers utilize quantum bits (qubits) with unique properties such as superposition and entanglement, enabling them to solve certain computational problems more efficiently and achieve significant speed-ups in specific applications.</div><div>In this manuscript, we evaluate the potential of quantum computing in agriculture and life sciences by reviewing computational challenges suitable for quantum computing and exploring exemplary domain applications. We examine optimization problems in agrifood supply chains, large-scale linear equation systems in animal breeding, quantum-based network architectures for machine learning in classifying satellite images for land-use analysis, quantum simulations for resource recovery from agriculture waste streams, and quantum search algorithms for genome assembly. Each computational problem type presents unique opportunities and challenges, underscoring the need for tailored quantum algorithms.</div><div>Furthermore, we provide a critical assessment of the broader potential of quantum computing, discussing its challenges, limitations, and how to facilitate a potential implementation. While current quantum hardware remains limited, developing quantum algorithms is still valuable — not only to prepare for future advancements but also to foster innovation through interdisciplinary collaboration. Rather than replacing traditional computing, we foresee quantum computing complementing classical systems, offering novel solutions to previously intractable problems. Continued research and interdisciplinary collaborations are essential to realize the full potential of quantum computing, paving the way for pioneering advancements in agriculture and life sciences.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110332"},"PeriodicalIF":7.7,"publicationDate":"2025-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143769015","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development of an ecohydrological model for coupled simulation of water and carbon fluxes, crop growth, and canopy spectra over croplands","authors":"Cheng Yang ,&nbsp;Huimin Lei ,&nbsp;Xingyu Hu ,&nbsp;Min Liu","doi":"10.1016/j.compag.2025.110336","DOIUrl":"10.1016/j.compag.2025.110336","url":null,"abstract":"<div><div>Canopy spectral information, such as Sun-Induced chlorophyll Fluorescence (SIF) and hyperspectral reflectance, are closely associated with photosynthesis and canopy structure. These spectral indicators provide valuable insights into the actual growth status of crops, thereby guiding management practices in agricultural ecosystems. While considerable efforts have been devoted to simulating the processes of photosynthesis and crop growth, comprehensive and mechanistic modeling of canopy spectral information, integrated with these processes, remains underexplored in traditional crop models. Considering the recent advances in remote sensing observations which are mostly emitted or reflected signals, being able to accurately reproduce the canopy spectra is also advantageous to enhancing the model applicability. In this study, we propose an ecohydrological model (namely the Weishan model) with an integration of a water-carbon-energy fluxes module, a carbon allocation module, a reflectance spectrum module, and a SIF spectrum module for both C₃ (winter wheat) and C₄ crops (summer maize). Comprehensive model calibration and validation have been conducted based on the eddy covariance observations over a typical winter wheat-summer maize rotation cropping cropland in the North China Plain. Validation results highlight the capability and applicability of our ecohydrological model in reproducing the variation of water-carbon fluxes (i.e., evaporation, transpiration, averaged soil moisture, and gross primary productivity), crop growth variables (i.e., leaf area index and end-of-season crop yield), and canopy spectral information (i.e., top-of-canopy SIF, reflectance at near-infrared, red, and blue bands, and vegetation indices). Our model is capable of simulating canopy spectra through mechanistic representations of photosynthesis (e.g., utilizing the Farquhar biochemical model, the Ball-Berry stomatal model, and the energy balance model) and crop dynamics (e.g., phenology, leaf dynamics, carbon allocation and partitioning, biomass accumulation, and yield formation). This comprehensive framework enables the model to effectively disentangle the complex interactions among these processes within a changing environmental context. Furthermore, the model’s ability to accurately reproduce canopy spectra highlights its potential to leverage remote sensing observations to enhance the model performance. We emphasize the functionality and future applicability of our model in advancing ecohydrological and agricultural research.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110336"},"PeriodicalIF":7.7,"publicationDate":"2025-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143769014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A geodesic distance regression-based semantic keypoints detection method for pig point clouds and body size measurement","authors":"Zhankang Xu ,&nbsp;Qifeng Li ,&nbsp;Weihong Ma ,&nbsp;Mingyu Li ,&nbsp;Daniel Morris ,&nbsp;Zhiyu Ren ,&nbsp;Chunjiang Zhao","doi":"10.1016/j.compag.2025.110285","DOIUrl":"10.1016/j.compag.2025.110285","url":null,"abstract":"<div><div>Pig body size reflects its physical shape and growth development, making accurate non-contact body size measurement crucial for practical farming production. The point cloud-based non-contact body size measurement method provides an effective alternative to traditional manual measurement, with the key challenge being the accurate identification of measurement keypoints. Many recent studies have focused solely on point cloud slicing or segmentation to indirectly locate body size keypoints, while research on directly predicting keypoints from livestock point clouds remains scarce. Therefore, we propose a method for directly detecting point clouds keypoints based on geodesic distance regression, which enables efficient measurement of pig body size through these keypoints. This approach transforms the detection of semantic keypoints in point clouds into a regression problem of geodesic distances between points and keypoints through heatmaps. The improved PointNet++ encoder-decoder architecture is utilized to learn distances on the manifold, enabling efficient keypoint detection. The model can be viewed as outputting probability values for each point corresponding to various keypoints, with the point having the highest probability selected as the predicted keypoint. Experimental results demonstrate an average root mean square error (RMSE) of 4.115 cm across eight keypoint types. The derived pig body size parameters achieve mean absolute percentage errors (MAPE) of 2.83 % for body length, 5.33 % for body width, 2.84 % for body height, 3.73 % for rump circumference, 4.83 % for thoracic circumference, and 3.83 % for abdominal circumference. The proposed geodesic distance regression-based semantic keypoints detection method for pig point clouds enables automated, accurate, and robust body size measurements, demonstrating significant potential for widespread application.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"234 ","pages":"Article 110285"},"PeriodicalIF":7.7,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143759685","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Detection of the number of wheat stems using multi-view images from smart glasses","authors":"Tao Liu ,&nbsp;Jianliang Wang ,&nbsp;Junfan Chen ,&nbsp;Weijun Zhang ,&nbsp;Ying Wang ,&nbsp;Yuanyuan Zhao ,&nbsp;Yi Sun ,&nbsp;Zhaosheng Yao ,&nbsp;Jiayi Wang ,&nbsp;Chengming Sun","doi":"10.1016/j.compag.2025.110370","DOIUrl":"10.1016/j.compag.2025.110370","url":null,"abstract":"<div><div>The number of stems in wheat populations is a fundamental parameter to achieve high yields and a critical agronomic trait in wheat production and variety selection. Although smart agricultural technology can estimate various agronomic parameters, the wheat stem is often obscured by multiple canopy leaves, making estimation challenging. Consequently, the current method to determine the stem number predominantly relies on labor-intensive manual techniques, which are inefficient and significantly influenced by subjective factors. This study proposes the use of augmented reality (AR) glasses as an imaging data acquisition tool to detect the number of wheat stems with high precision based on features from the top canopy and lateral images of wheat clusters. Following a correlation analysis, four color features, <em>Coverage</em>, the texture feature <em>Contrast</em>, and two lateral peak features SI (<em>Peaks1</em> and <em>Peaks2</em>) of the top canopy image were identified. The study comparatively analyzed the image features from three perspectives for their accuracy in detecting the number of wheat stems. The results indicated a strong correlation between the peak feature (SI) and the number of wheat stems with an <em>R²</em> value above 0.75. The estimation using only canopy image features (CC) resulted in significant errors, where the <em>RMSE</em> was 20 under high-density planting conditions. Using only <em>Peaks1</em> and <em>Peaks2</em> yielded higher accuracy in the stem estimation, but uncertainties persisted in some high-density scenarios. Furthermore, the study combined CC and SI for the estimation and used a random forest algorithm to construct a stem estimation model. This model maintained an <em>RMSE</em> below 10, even under high planting densities and below 5 under low densities, which demonstrated high accuracy. This study could provide insights into stem detection for crops similar to wheat and offer a reference for other studies that require hands-free and first-person perspective image acquisition.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110370"},"PeriodicalIF":7.7,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Research status of apple picking robotic arm picking strategy and end-effector","authors":"Chunlin Chen ,&nbsp;Zhuoying Song ,&nbsp;Xiang Li ,&nbsp;Chongcheng Chen ,&nbsp;Fuzeng Yang ,&nbsp;Zheng Wang","doi":"10.1016/j.compag.2025.110349","DOIUrl":"10.1016/j.compag.2025.110349","url":null,"abstract":"<div><div>Currently, the picking efficiency of apple picking robot is not high, easy to damage the fruit, which seriously affects the productization and marketization of apple picking robot. Therefore, it is necessary to conduct a systematic review of the research status of apple picking robots. Firstly, this paper summarizes the research status of apple picking robot. The picking motions that simulate human actions, obstacle avoidance strategies, trajectory planning algorithms, and multi-arm cooperative picking strategies are discussed and summarized in depth. Then, the key structure of the apple picking robot’s end-effector is analyzed. This study further elaborates on its picking mode, along with the rigidity and flexibility characteristics. Finally, aiming at the problems existing in robot picking strategies and picking end-effectors, this paper systematically reviews the research status and challenges in this field more comprehensively, and provides references for promoting the further development and application of apple picking robots.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110349"},"PeriodicalIF":7.7,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760780","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Detection and analysis of sow nursing behavior based on the number and location of piglets outside the suckling area using YOLOv5s","authors":"Luo Liu ,&nbsp;Jinxin Chen ,&nbsp;Qi-an Ding ,&nbsp;Ruqian Zhao ,&nbsp;Mingxia Shen ,&nbsp;Longshen Liu","doi":"10.1016/j.compag.2025.110324","DOIUrl":"10.1016/j.compag.2025.110324","url":null,"abstract":"<div><div>The nursing behavior of sows plays a crucial role in piglet growth, making precise monitoring and robust statistical analysis essential for a comprehensive evaluation of maternal characteristics. This study employed RGB cameras alongside an innovative Recognition Sow Nursing–You Only Look Once (RSN-YOLO) model to effectively monitor sow nursing behavior. The experimental comprised 24 sows of Yorkshire or Landrace breeds, each fostering no more than 13 piglets. The system automatically detected and recorded the start and end times of each nursing episode, enabling the collection and subsequent analysis of both individual sow characteristics and group behaviors. Results indicate that the YOLOv5s object detection model strikes an optimal balance between speed and accuracy, processing frames at 5.7 ms/frame, while achieving a precision rate of 96.3 %, a recall rate of 95.0 %, and a mean Average Precision ([email protected]) of 97.3 %. Comparisons between these automated detections and manual counts from continuous 24-hour video recordings across five pens confirmed that the method accurately captures both the number of nursing instances and the total duration of nursing with over 95 % accuracy when count errors do not exceed two occurrences. Even when count errors exceed two, accuracy remains above 92 %, with the average duration of each nursing session consistently measured with high precision. The study further revealed that sow nursing behavior does not exhibit a significant day-night rhythm, although notable individual variability within the group is evident. This variability is critical for early identification and intervention in cases where sows exhibit abnormal nursing behaviors relative to overall group patterns. Over a 21-day lactation period, both the total and daily average nursing durations decreased and subsequently stabilized as piglets aged, while the frequency of nursing remained relatively constant. A notable positive correlation (r = 0.67) was found between the number of nursing events and the total nursing duration. Additionally, the results support previous findings that proximally influences the synchronicity of nursing behavior: sows located farther apart are significantly less likely to nurse simultaneously (<em>P</em> &lt; 0.01). Overall, this methodology introduces a novel approach for automating the monitoring of sow nursing behavior on large-scale pig farms. By analyzing individual and group nursing patterns, the approach facilitates the early detection and warning of abnormal nursing behaviors, thereby enhancing the assessment of sow nursing performance and significantly advancing precision livestock farming.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110324"},"PeriodicalIF":7.7,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760851","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimization of motion strategy for a micro multi-functional chassis based on RBF neural network in intercropping mode","authors":"Hao Ling ,&nbsp;Tengfei Wu ,&nbsp;Yonghui Wu,&nbsp;Zheng Liu,&nbsp;Lihua Zhang,&nbsp;Xiaorong Lv","doi":"10.1016/j.compag.2025.110316","DOIUrl":"10.1016/j.compag.2025.110316","url":null,"abstract":"<div><div>The hilly and mountainous regions of China are characterized by unique features such as small plots of land, steep slopes, fragmented fields, and high soil viscosity, which result in a decline in the efficiency of conventional agricultural machinery, or even render its use impractical. To address this issue, this study developed a micro universal chassis adapted to hilly terrains. First, a four-wheel-drive multifunctional electric micro chassis was designed, considering the terrain characteristics of hilly regions and the agronomic requirements of maize-soybean strip intercropping. Second, the kinematics of the chassis were modeled and analyzed to determine optimal posture control strategies, and a fuzzy RBF neural network-based PID control algorithm was designed to enable dynamic adjustment of the chassis. Then, extensive testing was conducted on the prototype chassis, including straight-line driving tests, steering tests, climbing tests, and passability tests, which demonstrated its excellent operational performance. The straight-line driving tests showed an average lateral deviation of 30 mm and a maximum deviation of 60 mm, while the in-situ steering tests recorded a deviation of 20 mm. Finally, the prototype was applied to field weeding operations, where results indicated that its performance, including travel speed, weeding efficiency, and seedling damage rate, significantly outperformed existing traditional models. The findings suggest that the designed multifunctional micro universal chassis is highly effective for use in hilly and mountainous regions, with superior performance particularly under intercropping systems.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110316"},"PeriodicalIF":7.7,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Synthetic hyperspectral reflectance data augmentation by generative adversarial network to enhance grape maturity determination","authors":"Hongyi Lyu ,&nbsp;Miles Grafton ,&nbsp;Thiagarajah Ramilan ,&nbsp;Matthew Irwin ,&nbsp;Eduardo Sandoval","doi":"10.1016/j.compag.2025.110341","DOIUrl":"10.1016/j.compag.2025.110341","url":null,"abstract":"<div><div>Non-destructive and rapid grape maturity detection is important for the wine industry. The ongoing development of hyperspectral imaging techniques and deep learning methods has greatly helped in non-destructive assessing of grape quality and maturity, but the performance of deep learning methods depends on the volume and the quality of labeled data for training. Building non-destructive grape quality or maturity testing datasets requires damaging grapes for chemical analysis to produce labels which are time consuming and resource intensive. To solve this problem, this study proposed a conditional Wasserstain Generative Adversarial Network (WGAN) with the gradient penalty data augmentation technique to generate synthetic hyperspectral reflectance data of two grape maturity categories (ripe and unripe) and different Total Soluble Solids (TSS) values. The conditional WGAN with the gradient penalty was trained for a range of epochs: 500, 1000, 2000, 8000, 10,000, and 20,000. After training of 10,000 epochs, synthetic hyperspectral reflectance data were very similar to real spectra for each maturity category and different TSS values. Thereafter, contextual deep three-dimensional CNN (3D-CNN), Spatial Residual Network (SSRN) and Support Vector Machine (SVM) are trained on original training and synthetic + original training datasets to classify grape maturity. The synthetic hyperspectral reflectance data, incrementally added to the original training set in steps of 250, 500, 1000, 1500, and 2000 samples, consistently resulted in higher model performance compared to training solely on the original dataset. The best results were achieved by augmenting the training dataset with 2000 synthetic samples and training with a 3D-CNN, yielding a classification accuracy of 91 % on the testing set. To better assess the effectiveness of GAN-based data augmentation methods, two widely used regression models: Partial Least Squares Regression (PLSR) and one-dimensional CNN (1D-CNN) were used based on same data augmentation method. The best result was achieved by adding 250 synthetic samples to the original training set when training 1D-CNN model, yielding an R² of 0.78, RMSE of 0.63 °Brix, and RPIQ of 3.36 on the testing set. This study indicated that deep learning models combined with conditional WGAN with the gradient penalty data augmentation technique had a good application prospect in the grape maturity assessment.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110341"},"PeriodicalIF":7.7,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Drone multispectral imaging captures the effects of soil mineral nitrogen on canopy structure and nitrogen use efficiency in wheat","authors":"Jie Wang ,&nbsp;Sebastian T. Meyer ,&nbsp;Xijie Xu ,&nbsp;Wolfgang W. Weisser ,&nbsp;Kang Yu","doi":"10.1016/j.compag.2025.110342","DOIUrl":"10.1016/j.compag.2025.110342","url":null,"abstract":"<div><div>Drone remote sensing offers a powerful tool for monitoring vegetation and agricultural systems. However, its effectiveness in assessing the effect of soil mineral nitrogen (<em>N</em>_min) on crop canopy traits remains inadequately explored. This study investigates the relationship between soil <em>N</em>_min variability and canopy characteristics, grain yield, and nitrogen use efficiency (NUE), and explores the potential to predict NUE using drone multispectral images. Multispectral data were collected across growth stages over two growing seasons. The analysis revealed that soil <em>N</em>_min significantly affected canopy structure, with low <em>N</em>_min inducing a ’blue shift’ of the red-edge spectral position. The multilayer perceptron regression model predicted NUE with high accuracy (R² &gt; 0.7) in early growth stages, identifying red-edge spectral indices and canopy height as key predictors. Texture features did not play a significant role in the models for predicting NUE, which remains to be further understood in future research. These findings highlight the capability of UAV remote sensing data, especially the red-edge spectral features, to capture the effects of soil <em>N</em>_min on canopy traits. This study provides a proof-of-concept for mapping NUE using UAV images, with the final goal of improving crop nitrogen management and fertilizer use efficiency in agriculture.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110342"},"PeriodicalIF":7.7,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760850","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Improving chili pepper LAI prediction with TPE-2BVIs and UAV hyperspectral imagery","authors":"Haiyang Zhang,&nbsp;Guolong Wang,&nbsp;Fanfan Song,&nbsp;Zhaoqi Wen,&nbsp;Wenwen Li,&nbsp;Ling Tong,&nbsp;Shaozhong Kang","doi":"10.1016/j.compag.2025.110368","DOIUrl":"10.1016/j.compag.2025.110368","url":null,"abstract":"<div><div>Leaf area index (LAI) of chili peppers is an important indicator of plant growth and productivity. Accurate monitoring of LAI is crucial for optimizing growth conditions, and improving crop yields. However, existing remote sensing-based methods for LAI prediction often face low accuracy, especially in high-density vegetation and across different growth stages. These limitations primarily stem from issues such as multicollinearity and the constraints of spectral information processing mechanisms, which affect the stability and accuracy of predictions. To address these challenges, this study proposed a novel prediction method combining a tree-structured parzen estimator (TPE)-optimized two-band vegetation indices (TPE-2BVIs) with random forest regression (RFR). The TPE-2BVIs band optimization algorithm effectively extracts latent spectral information, enhancing LAI prediction stability and accuracy by addressing multicollinearity and nonlinear model complexity. Experimental results show that: (1) hyperspectral images provide more detailed spectral information compared to multispectral images, significantly improving the accuracy of LAI prediction; (2) the use of 2BVIs alleviates the effects of spectral saturation and dynamic variations during different growth stages, further improving prediction accuracy; (3) the proposed TPE-2BVIs band optimization method significantly enhances both the performance and stability of the model. When combined with RFR, the model achieves <em>R</em>² = 0.887, RMSE = 0.520, and NRMSE = 7.554 %. The TPE-2BVIs band optimization algorithm introduced in this study effectively extracts latent spectral information, overcoming the limitations of multicollinearity and the complexity of nonlinear models in traditional methods. This approach significantly improves the stability and accuracy of LAI predictions. The proposed method provides an innovative solution for remote sensing vegetation monitoring and agricultural applications, offering broad potential for estimating phenotypic parameters under diverse environmental conditions.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"235 ","pages":"Article 110368"},"PeriodicalIF":7.7,"publicationDate":"2025-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143747601","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}