Pub Date : 2025-12-01DOI: 10.1016/j.inpa.2025.07.003
Helong Yu , Liyun Han , Chengcheng Chen , Honghong Su , Qichao Niu , Ronghao Meng , Mingxuan Xue
The accurate counting of overlapping watermelon seeds is a key foundation for seed quality testing, breeding selection, resource allocation, and other processes. To improve the counting accuracy for flat and slightly overlapping seeds, we introduce LOYOLO-GC, a Lightweight Occlusion YOLO8n-based group counting model. It adopts HGNetV2 as its backbone, where HGBlocks extract multi-level features for improved learning. GhostConv replaces the standard convolution in HGBlocks, forming LightHGBlock to reduce the number of parameters by generating intrinsic and ghost feature maps with fewer kernels. In addition, a Large Separable Kernel Attention mechanism (LSKA) is used to decompose deep convolution kernels into horizontal and vertical 1D kernels, enabling efficient large kernel attention with lower computational and memory cost. After optimizing the model, we build a multi-occlusion watermelon seed dataset and employ it to develop a LOYOLO-based group counting method. The experimental results show that LOYOLO-GC outperforms SOTA models, achieving 96.08 % accuracy and 86.66 % mAP, an improvement of 0.48 % and 1.67 %, respectively. The model parameters decrease by 63.8 % and GMACs decrease by 38.9 %. Counting accuracy is also improved, with ACC increasing by 5.32 % and L-ACC increasing by 5.04 %, while MAE and RMSE are decreased by 3.68 and 3.28, respectively.
{"title":"Lightweight precision model for watermelon seed group density estimation and counting","authors":"Helong Yu , Liyun Han , Chengcheng Chen , Honghong Su , Qichao Niu , Ronghao Meng , Mingxuan Xue","doi":"10.1016/j.inpa.2025.07.003","DOIUrl":"10.1016/j.inpa.2025.07.003","url":null,"abstract":"<div><div>The accurate counting of overlapping watermelon seeds is a key foundation for seed quality testing, breeding selection, resource allocation, and other processes. To improve the counting accuracy for flat and slightly overlapping seeds, we introduce LOYOLO-GC, a Lightweight Occlusion YOLO8n-based group counting model. It adopts HGNetV2 as its backbone, where HGBlocks extract multi-level features for improved learning. GhostConv replaces the standard convolution in HGBlocks, forming LightHGBlock to reduce the number of parameters by generating intrinsic and ghost feature maps with fewer kernels. In addition, a Large Separable Kernel Attention mechanism (LSKA) is used to decompose deep convolution kernels into horizontal and vertical 1D kernels, enabling efficient large kernel attention with lower computational and memory cost. After optimizing the model, we build a multi-occlusion watermelon seed dataset and employ it to develop a LOYOLO-based group counting method. The experimental results show that LOYOLO-GC outperforms SOTA models, achieving 96.08 % accuracy and 86.66 % mAP, an improvement of 0.48 % and 1.67 %, respectively. The model parameters decrease by 63.8 % and GMACs decrease by 38.9 %. Counting accuracy is also improved, with ACC increasing by 5.32 % and L-ACC increasing by 5.04 %, while MAE and RMSE are decreased by 3.68 and 3.28, respectively.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 4","pages":"Pages 565-580"},"PeriodicalIF":7.4,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145697905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The accurate recognition of apple leaf diseases is crucial for ensuring crop health and agricultural productivity. However, deep learning models often suffer from poor generalization across diverse environments due to variations in lighting, background complexity, and leaf appearance. To address these challenges, we proposed EConv-ViT, a novel robust generalization model integrating ConvNeXt and Vision Transformer (ViT), enhanced with Efficient Channel Attention (ECA) for superior feature extraction and DropKey to improve generalization and applied the mode on image dataset both captured in laboratory and natural environments for healthy apple leaves, alternaria blotch, grey spot, rust, and mosaic disease. The propsed EConv-ViT model was tested on an independent dataset and achieved accuracy of 99.2% on laboratory-captured image dataset and 79.3% on images captured in natural environments. The classification accuracy for EConv-ViT model exhibited 18.6%, 36.1% and 37.8% improvements compared with ViT, ConvNeXt, and ResNet50 models on a dataset captured in natural environments. EConv-ViT can effectively capture both local and global features and demonstrate its potential for the application on related automated disease monitoring systems.
{"title":"EConv-ViT: A strongly generalized apple leaf disease classification model based on the fusion of ConvNeXt and Transformer","authors":"Xin Huang , Demin Xu , Yongqiao Chen , Qian Zhang , Puyu Feng , Yuntao Ma , Qiaoxue Dong , Feng Yu","doi":"10.1016/j.inpa.2025.03.001","DOIUrl":"10.1016/j.inpa.2025.03.001","url":null,"abstract":"<div><div>The accurate recognition of apple leaf diseases is crucial for ensuring crop health and agricultural productivity. However, deep learning models often suffer from poor generalization across diverse environments due to variations in lighting, background complexity, and leaf appearance. To address these challenges, we proposed EConv-ViT, a novel robust generalization model integrating ConvNeXt and Vision Transformer (ViT), enhanced with Efficient Channel Attention (ECA) for superior feature extraction and DropKey to improve generalization and applied the mode on image dataset both captured in laboratory and natural environments for healthy apple leaves, alternaria blotch, grey spot, rust, and mosaic disease. The propsed EConv-ViT model was tested on an independent dataset and achieved accuracy of 99.2% on laboratory-captured image dataset and 79.3% on images captured in natural environments. The classification accuracy for EConv-ViT model exhibited 18.6%, 36.1% and 37.8% improvements compared with ViT, ConvNeXt, and ResNet50 models on a dataset captured in natural environments. EConv-ViT can effectively capture both local and global features and demonstrate its potential for the application on related automated disease monitoring systems.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 4","pages":"Pages 466-477"},"PeriodicalIF":7.4,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145697844","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.inpa.2025.03.002
Weizhong Yu , Yizheng Huang , Fang Ji , Yude Yu , Dongxian He , Zhao Li
Microbial contamination is an inevitable challenge in plant factory, posing substantial risks of economic loss and potential threats to human health if not addressed promptly. However, existing detection methods are characterized by prolonged processing times, high costs, and dependence on skilled technicians, limiting their practicality for routine monitoring. Therefore, there is a critical need for the development of rapid, cost-effective, and reliable device for the quantitative monitoring of microorganisms in both the air and nutrient solutions of the plant factory. We have developed an integrated microfluidic biosensor that can be used to quantitatively monitor microbial levels in air and nutrient solutions by combining ATP bioluminescence. The biosensor was verified and calibrated through a standard ATP solution with Bacillus subtilis bacterial solution, followed by testing of the real air and nutrient solution samples from plant factories. The detection process on the microfluidic chip was automatically controlled to complete within 3 min. The consumption of ATP reaction solution and lysate for one assay was about 10 μL and 16 μL, respectively. The sensitivity of bacterial quantification was up to 6.4 × 103 CFU mL−1 with a detection range covering 4 orders of magnitude. This biosensor has been demonstrated to have similar detection accuracy with the culture counting method and enable quantitative monitoring of microorganisms in plant factory, while greatly reducing the detection cycles.
{"title":"A microfluidic biosensor for microbial quantitative monitoring of air and nutrient solution in the plant factory","authors":"Weizhong Yu , Yizheng Huang , Fang Ji , Yude Yu , Dongxian He , Zhao Li","doi":"10.1016/j.inpa.2025.03.002","DOIUrl":"10.1016/j.inpa.2025.03.002","url":null,"abstract":"<div><div>Microbial contamination is an inevitable challenge in plant factory, posing substantial risks of economic loss and potential threats to human health if not addressed promptly. However, existing detection methods are characterized by prolonged processing times, high costs, and dependence on skilled technicians, limiting their practicality for routine monitoring. Therefore, there is a critical need for the development of rapid, cost-effective, and reliable device for the quantitative monitoring of microorganisms in both the air and nutrient solutions of the plant factory. We have developed an integrated microfluidic biosensor that can be used to quantitatively monitor microbial levels in air and nutrient solutions by combining ATP bioluminescence. The biosensor was verified and calibrated through a standard ATP solution with <em>Bacillus subtilis</em> bacterial solution, followed by testing of the real air and nutrient solution samples from plant factories. The detection process on the microfluidic chip was automatically controlled to complete within 3 min. The consumption of ATP reaction solution and lysate for one assay was about 10 μL and 16 μL, respectively. The sensitivity of bacterial quantification was up to 6.4 × 10<sup>3</sup> CFU mL<sup>−1</sup> with a detection range covering 4 orders of magnitude. This biosensor has been demonstrated to have similar detection accuracy with the culture counting method and enable quantitative monitoring of microorganisms in plant factory, while greatly reducing the detection cycles.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 4","pages":"Pages 478-486"},"PeriodicalIF":7.4,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145697845","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.inpa.2025.09.004
Gedi Liu , Keyang Zhong , Huilin Li , Tao Chen , Yang Wang
{"title":"Corrigendum to “A state of art review on time series forecasting with machine learning for environmental parameters in agricultural greenhouses” [Inf. Process. Agric. 11(2) (2024) 143–162]","authors":"Gedi Liu , Keyang Zhong , Huilin Li , Tao Chen , Yang Wang","doi":"10.1016/j.inpa.2025.09.004","DOIUrl":"10.1016/j.inpa.2025.09.004","url":null,"abstract":"","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 4","pages":"Page 595"},"PeriodicalIF":7.4,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145697951","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-01DOI: 10.1016/j.inpa.2024.10.002
Raul Toscano-Miranda , Jose Aguilar , Manuel Caro , Anibal Trebilcok , Mauricio Toro
Precision farming (PF) allows the efficient use of resources such as water, and fertilizers, among others; as well, it helps to analyze the behavior of insect pests, in order to increase production and decrease the cost of crop management. This paper introduces an innovative approach to integrated cotton management, involving the implementation of an Autonomous Cycle of Data Analysis Tasks (ACODAT). The proposed autonomous cycle is composed of a classification task of the population of pests (boll weevil) (based on eXtreme Gradient Boosting-XGBoost), a diagnosis-prediction task of cotton yield (based on a fuzzy system), and a prescription task of strategies for the adequate management of the crop (based on genetic algorithms). The proposed system can evaluate several variables according to the conditions of the crop, and recommend the best strategy for increasing the cotton yield. In particular, the classification task has an accuracy of 88%, the diagnosis/prediction task obtained an accuracy of 98 %, and the genetic algorithm recommends the best strategy for the context analyzed. Focused on integrated cotton management, our system offers flexibility and adaptability, which facilitates the incorporation of new tasks.
{"title":"Precision farming using autonomous data analysis cycles for integrated cotton management","authors":"Raul Toscano-Miranda , Jose Aguilar , Manuel Caro , Anibal Trebilcok , Mauricio Toro","doi":"10.1016/j.inpa.2024.10.002","DOIUrl":"10.1016/j.inpa.2024.10.002","url":null,"abstract":"<div><div>Precision farming (PF) allows the efficient use of resources such as water, and fertilizers, among others; as well, it helps to analyze the behavior of insect pests, in order to increase production and decrease the cost of crop management. This paper introduces an innovative approach to integrated cotton management, involving the implementation of an Autonomous Cycle of Data Analysis Tasks (ACODAT). The proposed autonomous cycle is composed of a classification task of the population of pests (boll weevil) (based on eXtreme Gradient Boosting-XGBoost), a diagnosis-prediction task of cotton yield (based on a fuzzy system), and a prescription task of strategies for the adequate management of the crop (based on genetic algorithms). The proposed system can evaluate several variables according to the conditions of the crop, and recommend the best strategy for increasing the cotton yield. In particular, the classification task has an accuracy of 88%, the diagnosis/prediction task obtained an accuracy of 98 %, and the genetic algorithm recommends the best strategy for the context analyzed. Focused on integrated cotton management, our system offers flexibility and adaptability, which facilitates the incorporation of new tasks.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 3","pages":"Pages 326-343"},"PeriodicalIF":7.4,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145327144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-01DOI: 10.1016/j.inpa.2024.09.007
Susanto B. Sulistyo , Arief Sudarmaji , Pepita Haryanti , Purwoko H. Kuncoro
Granulated coconut sugar has been well-known as a sweetener which is more nutritious and has lower glycemic index than cane sugar. Adding cane sugar to coconut sap during heating may result in coconut sugar with an undesirable export quality. The purpose of this study was to develop a novel approach by designing a low-cost portable spectrometer capable of detecting the presence of cane sugar in granulated coconut sugar using machine learning. The AS7265x multispectral sensor chipset is the main component of the proposed LED-based spectrometer. This chipset uses two integrated LEDs as the light source and has 18 channels output ranging from the visible to near-infrared spectrum as the predictor variables to identify the adulteration in granulated coconut sugar. A variety of machine learning techniques were used to determine the purity of granulated coconut sugar as well as the quantity of cane sugar added. Backpropagation neural networks outperformed various machine learning methods, including the support vector machine, k-nearest neighbor, and naïve Bayes methods, in determining the purity of granulated coconut sugar. The developed portable LED-based spectrometer by means of backpropagation neural networks as the classifier can successfully detect adulteration in granulated coconut sugar with very high accuracy level.
{"title":"A novel approach for detection of granulated coconut sugar adulteration using LED-based spectrometer and machine learning","authors":"Susanto B. Sulistyo , Arief Sudarmaji , Pepita Haryanti , Purwoko H. Kuncoro","doi":"10.1016/j.inpa.2024.09.007","DOIUrl":"10.1016/j.inpa.2024.09.007","url":null,"abstract":"<div><div>Granulated coconut sugar has been well-known as a sweetener which is more nutritious and has lower glycemic index than cane sugar. Adding cane sugar to coconut sap during heating may result in coconut sugar with an undesirable export quality. The purpose of this study was to develop a novel approach by designing a low-cost portable spectrometer capable of detecting the presence of cane sugar in granulated coconut sugar using machine learning. The AS7265x multispectral sensor chipset is the main component of the proposed LED-based spectrometer. This chipset uses two integrated LEDs as the light source and has 18 channels output ranging from the visible to near-infrared spectrum as the predictor variables to identify the adulteration in granulated coconut sugar. A variety of machine learning techniques were used to determine the purity of granulated coconut sugar as well as the quantity of cane sugar added. Backpropagation neural networks outperformed various machine learning methods, including the support vector machine, k-nearest neighbor, and naïve Bayes methods, in determining the purity of granulated coconut sugar. The developed portable LED-based spectrometer by means of backpropagation neural networks as the classifier can successfully detect adulteration in granulated coconut sugar with very high accuracy level.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 3","pages":"Pages 300-311"},"PeriodicalIF":7.4,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145327147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-01DOI: 10.1016/j.inpa.2024.12.002
Bernard Roger Ramos Collin , Danilo de Lima Alves Xavier , Thiago Magalhães Amaral , Ana Cristina G. Castro Silva , Daniel dos Santos Costa , Fernanda Magalhães Amaral , Jefferson Tales Oliva
The importance of identifying the caliber in advance is in knowing the exact quantity of mangos, by weight, that a determined crop season (complete periods of the mango cycle from growth up to fruit harvest) will provide. This study uses Random Forest method to predict the percentage distribution of the calibers of four mango varieties from Brazil’s largest exporter and producer. Our proposed approach was conducted in the following steps: data collection; data preprocessing; predictive model building; and model evaluation. The data correspond to three crop seasons, namely those of 2019, 2020, and 2021. Each data line corresponds to a plot with the percentage of a determined caliber at the end of a crop season. The number of rows in the dataset is 5503, with 37.33 %, 31.47 %, 22.76 %, and 8.44 % corresponding to the Keitt, Tommy Atkins, Kent, and Palmer varieties, respectively. The variables are Productivity, (N) Nitrogen, Number of plants (units), Plants/hectare, Month of floral induction, (Zn) Zinc, (S) Sulfur, (B) Boron, Caliber, and Percentage of caliber. The Python programming language was used to preprocess the data, do exploratory analysis, develop the algorithms of the Random Forest Regressor, and compile the lines of the code in Visual Studio Code. Python libraries were used during the study, such as pandas for data handling and Scipy for removing outliers to avoid any biases in the data. The YellowBrick library was used for the feature selection process. Four regression models were created using Random Forest (RF), one for each variety of fruit that composes the dataset. The algorithms showed satisfactory results for Kent, Keitt, Tommy Atkins, and Palmer mangoes, with the following R2 of the models: 87.29 %, 74.37 %, 87.69 %, and 62.75 %, respectively. During the Feature Selection step, nitrogen (N) was perceived to be highly important in all the models, highlighting the representative nature of this element in fruit formation. From the models created, it is possible to predict the percentage distribution of the calibers of mangos from each growing area 6 months in advance, using data that characterize each area and information on the presence of leaf nutrients as input.
提前确定口径的重要性在于知道芒果的确切数量（按重量计算），这是一个确定的作物季节（芒果从生长到收获的完整周期）将提供的。本研究使用随机森林方法来预测来自巴西最大出口国和生产国的四种芒果品种的直径百分比分布。我们提出的方法分为以下几个步骤：数据收集；数据预处理;预测模型构建；以及模型评估。数据对应三个作物季节，即2019年、2020年和2021年。每条数据线对应一个地块，在作物季节结束时确定口径的百分比。数据集中的行数为5503，分别对应于Keitt、Tommy Atkins、Kent和Palmer品种的行数分别为37.33 %、31.47 %、22.76 %和8.44 %。变量为生产力、(N)氮、株数（单位）、株数/公顷、诱导花月、（Zn）锌、(S)硫、(B)硼、口径和口径百分比。使用Python编程语言对数据进行预处理，进行探索性分析，开发随机森林回归器的算法，并在Visual Studio code中编译代码行。在研究过程中使用了Python库，例如pandas用于数据处理，Scipy用于去除异常值以避免数据中的任何偏差。在特性选择过程中使用了YellowBrick库。使用随机森林（RF）创建了四个回归模型，每个模型对应组成数据集的水果品种。对于Kent， Keitt， Tommy Atkins和Palmer芒果，算法显示了令人满意的结果，模型的R2分别为87.29 %,74.37 %，87.69 %和62.75 %。在特征选择步骤中，氮(N)在所有模型中都被认为是非常重要的，突出了该元素在果实形成中的代表性。根据所创建的模型，可以提前6 个月预测每个种植区域芒果直径的百分比分布，使用每个区域的特征数据和叶片营养成分的存在信息作为输入。
{"title":"Random forest regressor applied in prediction of percentages of calibers in mango production","authors":"Bernard Roger Ramos Collin , Danilo de Lima Alves Xavier , Thiago Magalhães Amaral , Ana Cristina G. Castro Silva , Daniel dos Santos Costa , Fernanda Magalhães Amaral , Jefferson Tales Oliva","doi":"10.1016/j.inpa.2024.12.002","DOIUrl":"10.1016/j.inpa.2024.12.002","url":null,"abstract":"<div><div>The importance of identifying the caliber in advance is in knowing the exact quantity of mangos, by weight, that a determined crop season (complete periods of the mango cycle from growth up to fruit harvest) will provide. This study uses Random Forest method to predict the percentage distribution of the calibers of four mango varieties from Brazil’s largest exporter and producer. Our proposed approach was conducted in the following steps: data collection; data preprocessing; predictive model building; and model evaluation. The data correspond to three crop seasons, namely those of 2019, 2020, and 2021. Each data line corresponds to a plot with the percentage of a determined caliber at the end of a crop season. The number of rows in the dataset is 5503, with 37.33 %, 31.47 %, 22.76 %, and 8.44 % corresponding to the Keitt, Tommy Atkins, Kent, and Palmer varieties, respectively. The variables are Productivity, (N) Nitrogen, Number of plants (units), Plants/hectare, Month of floral induction, (Zn) Zinc, (S) Sulfur, (B) Boron, Caliber, and Percentage of caliber. The Python programming language was used to preprocess the data, do exploratory analysis, develop the algorithms of the Random Forest Regressor, and compile the lines of the code in Visual Studio Code. Python libraries were used during the study, such as pandas for data handling and Scipy for removing outliers to avoid any biases in the data. The YellowBrick library was used for the feature selection process. Four regression models were created using Random Forest (RF), one for each variety of fruit that composes the dataset. The algorithms showed satisfactory results for Kent, Keitt, Tommy Atkins, and Palmer mangoes, with the following R<sup>2</sup> of the models: 87.29 %, 74.37 %, 87.69 %, and 62.75 %, respectively. During the Feature Selection<!--> <!-->step, nitrogen (N) was perceived to be highly important in all the models, highlighting the representative nature of this element in fruit formation. From the models created, it is possible to predict the percentage distribution of the calibers of mangos from each growing area 6 months in advance, using data that characterize each area and information on the presence of leaf nutrients as input.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 3","pages":"Pages 370-383"},"PeriodicalIF":7.4,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145327149","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-01DOI: 10.1016/j.inpa.2024.11.001
Dan Xu , Lei Xu , Shusheng Wang , Mingqin Wang , Juncheng Ma , Chen Shi
Maximizing profit is usually the objective of optimal control of greenhouse cultivation. However, due to the problem of “the curse of dimensionality”, the global optimization of greenhouse climate is usually difficult when faced with a complex dynamic model and a long cultivation period. Compared with leafy vegetables with a much simpler dynamic model and a much shorter cultivation period, the year-round tomato model usually has many more states to describe its dynamics better. To solve the year-round climate control of greenhouse tomato cultivation, a rule-based model predictive control (MPC) algorithm is raised. The innovation of this paper lies in that the setpoints of the proposed MPC algorithms are determined by the external weather and the month-averaged predictions of the tomato price. With the greenhouse climate – tomato growth dynamic model and the economic performance index, different MPC algorithms are compared with the traditional on/off control algorithm and the open field cultivation. Quantified results of yield, cost, and profit are obtained with the weather data and market data collected in Beijing. Findings of this paper showed that the year-round greenhouse tomato cultivation in Beijing is hardly profitable with the tomato price sold as an open field product (XFD price). With the tomato price sold as a high-tech greenhouse product (JD price), the higher yield guarantees a higher profit. Moreover, the simple emphasis on energy minimization cannot even guarantee a higher yield than that in the open field. A synthetical consideration of yield and cost is a prerequisite for a high profit.
{"title":"Rule-based year-round model predictive control of greenhouse tomato cultivation: A simulation study","authors":"Dan Xu , Lei Xu , Shusheng Wang , Mingqin Wang , Juncheng Ma , Chen Shi","doi":"10.1016/j.inpa.2024.11.001","DOIUrl":"10.1016/j.inpa.2024.11.001","url":null,"abstract":"<div><div>Maximizing profit is usually the objective of optimal control of greenhouse cultivation. However, due to the problem of “the curse of dimensionality”, the global optimization of greenhouse climate is usually difficult when faced with a complex dynamic model and a long cultivation period. Compared with leafy vegetables with a much simpler dynamic model and a much shorter cultivation period, the year-round tomato model usually has many more states to describe its dynamics better. To solve the year-round climate control of greenhouse tomato cultivation, a rule-based model predictive control (MPC) algorithm is raised. The innovation of this paper lies in that the setpoints of the proposed MPC algorithms are determined by the external weather and the month-averaged predictions of the tomato price. With the greenhouse climate – tomato growth dynamic model and the economic performance index, different MPC algorithms are compared with the traditional on/off control algorithm and the open field cultivation. Quantified results of yield, cost, and profit are obtained with the weather data and market data collected in Beijing. Findings of this paper showed that the year-round greenhouse tomato cultivation in Beijing is hardly profitable with the tomato price sold as an open field product (XFD price). With the tomato price sold as a high-tech greenhouse product (JD price), the higher yield guarantees a higher profit. Moreover, the simple emphasis on energy minimization cannot even guarantee a higher yield than that in the open field. A synthetical consideration of yield and cost is a prerequisite for a high profit.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 3","pages":"Pages 344-357"},"PeriodicalIF":7.4,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145327413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-01DOI: 10.1016/j.inpa.2025.02.001
Qingwei Meng , Wei Qi Yan , Cong Xu , Zhaoxu Zhang , Xia Hao , Hui Chen , Wei Liu , Yanjie Li
Quantifying the leaf density and coloration of trees is critical for assessing landscape esthetics and photosynthetic efficiency; however, traditional leaf-counting methods are labor-intensive and potentially harmful to trees, making accurate measurements challenging. To address these issues, we present “Sassafras-net,” an advanced model specifically designed to detect and count colored leaves on Sassafras tzumu trees.
The methodology consists of two steps. First, we used an improved model termed YOLOX-CBAM to accurately detect and isolate individual trees. This model proved to be more effective than alternatives, such as YOLOX, YOLOv8, YOLOv7, YOLOv5, and Fater-RCNN. Second, the Sassafras-net model, which is based on the CCTrans network, counts the number of colored leaves per tree. Compared with the original CCTrans model of 52.30 and 84.90, the Sassafras-net model achieved significantly lower mean absolute error and mean squared error values of 27.29 and 39.00, respectively. These results confirm the ability of the model to accurately and efficiently quantify colored leaves.
To the best of our knowledge, this is the first study to quantify colored leaves in trees. Our method provides forestry researchers with an effective and economical tool for selecting and breeding S. tzumu trees with enhanced color traits. In addition, this study opens new avenues for studying tree traits related to leaf coloration.
{"title":"Optimization of Sassafras tzumu leaves color quantification with UAV RGB imaging and Sassafras-net","authors":"Qingwei Meng , Wei Qi Yan , Cong Xu , Zhaoxu Zhang , Xia Hao , Hui Chen , Wei Liu , Yanjie Li","doi":"10.1016/j.inpa.2025.02.001","DOIUrl":"10.1016/j.inpa.2025.02.001","url":null,"abstract":"<div><div>Quantifying the leaf density and coloration of trees is critical for assessing landscape esthetics and photosynthetic efficiency; however, traditional leaf-counting methods are labor-intensive and potentially harmful to trees, making accurate measurements challenging. To address these issues, we present “Sassafras-net,” an advanced model specifically designed to detect and count colored leaves on <em>Sassafras tzumu</em> trees.</div><div>The methodology consists of two steps. First, we used an improved model termed YOLOX-CBAM to accurately detect and isolate individual trees. This model proved to be more effective than alternatives, such as YOLOX, YOLOv8, YOLOv7, YOLOv5, and Fater-RCNN. Second, the Sassafras-net model, which is based on the CCTrans network, counts the number of colored leaves per tree. Compared with the original CCTrans model of 52.30 and 84.90, the Sassafras-net model achieved significantly lower mean absolute error and mean squared error values of 27.29 and 39.00, respectively. These results confirm the ability of the model to accurately and efficiently quantify colored leaves.</div><div>To the best of our knowledge, this is the first study to quantify colored leaves in trees. Our method provides forestry researchers with an effective and economical tool for selecting and breeding <em>S. tzumu</em> trees with enhanced color traits. In addition, this study opens new avenues for studying tree traits related to leaf coloration.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 3","pages":"Pages 384-397"},"PeriodicalIF":7.4,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145327416","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-01DOI: 10.1016/j.inpa.2025.02.002
Jin-Hai Li , Lie-Fei Ma , Wei-Wei Zhang , Ai-Li Qu , Yao-Yao Gao , De-Hua Gao , Yu-Tan Wang
Fresh Lycium barbarum L. (L. barbarum) fruits are renowned for their exceptionally high nutritional value and health benefits, which is leading to an increasing demand among consumers. However, the quality testing and grading of fresh L. barbarum fruits present significant challenges that hinder the growth of the L. barbarum industry. In this study, an electrical characterization method is used to analyze the variations in electrical parameters of fresh L. barbarum fruits under different degrees of damage. Optimal testing conditions for eight electrical parameters are determined, and principal component analysis (PCA) along with partial least squares (PLS) is applied to reduce data dimensionality and extract key features. Subsequently, damage degree discrimination models are developed using the support vector machine (SVM), random forest (RF), and convolutional neural network (CNN). The experimental results indicate that the PLS-RF model was the most effective, achieving discrimination accuracies of 99.48% and 91.25% in the training and test sets, respectively. The aim of this study is to validate the feasibility of using electrical characteristics to differentiate the degree of fruit damage and it establishes a reliable model for assessing damage extent in L. barbarum fruits. This innovative approach not only provides a novel method for evaluating fruit damage but may also serve as a theoretical basis for the development of mechanical harvesting equipment for L. barbarum fruits.
{"title":"Grading the damage degree of fresh Lycium barbarum L. fruits based on electrical characteristics","authors":"Jin-Hai Li , Lie-Fei Ma , Wei-Wei Zhang , Ai-Li Qu , Yao-Yao Gao , De-Hua Gao , Yu-Tan Wang","doi":"10.1016/j.inpa.2025.02.002","DOIUrl":"10.1016/j.inpa.2025.02.002","url":null,"abstract":"<div><div>Fresh <em>Lycium barbarum L.</em> (<em>L. barbarum</em>) fruits are renowned for their exceptionally high nutritional value and health benefits, which is leading to an increasing demand among consumers. However, the quality testing and grading of fresh <em>L. barbarum</em> fruits present significant challenges that hinder the growth of the <em>L. barbarum</em> industry. In this study, an electrical characterization method is used to analyze the variations in electrical parameters of fresh <em>L. barbarum</em> fruits under different degrees of damage. Optimal testing conditions for eight electrical parameters are determined, and principal component analysis (PCA) along with partial least squares (PLS) is applied to reduce data dimensionality and extract key features. Subsequently, damage degree discrimination models are developed using the support vector machine (SVM), random forest (RF), and convolutional neural network (CNN). The experimental results indicate that the PLS-RF model was the most effective, achieving discrimination accuracies of 99.48% and 91.25% in the training and test sets, respectively. The aim of this study is to validate the feasibility of using electrical characteristics to differentiate the degree of fruit damage and it establishes a reliable model for assessing damage extent in <em>L. barbarum</em> fruits. This innovative approach not only provides a novel method for evaluating fruit damage but may also serve as a theoretical basis for the development of mechanical harvesting equipment for <em>L. barbarum</em> fruits.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 3","pages":"Pages 398-407"},"PeriodicalIF":7.4,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145327414","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号