检索结果-内蒙古大学图书馆

Diagnosis of leaf chlorophyll content based on close-range multispectral fluorescence image correction

COMPUTERS AND ELECTRONICS IN agrICULTURE 2025年 231卷

作者： Liu, Guohui Liu, Mingjia Chao, Jinyang Tang, Weijie Zhao, Ruomei Liu, Yang Sun, Hong China Agr Univ Key Lab Smart Agr Syst Minist Educ Beijing 100083 Peoples R China China Agr Univ Key Lab Agr Informat Acquisit Technol Minist Agr Beijing 100083 Peoples R China

Multispectral fluorescence imaging is an effective tool for studying plant stress responses and diagnosing nutrient deficiencies. To address the influence of fluorescence shadow error caused by maize leaf structure and excitation light on chlorophyll content diagnosis, the BLF-CLAHE (Butterworth Low Filter Contrast Limited Adaptive Histogram Equalization) method was used to enhance the image containing fluorescence shadow to optimize the diagnosis of maize chlorophyll content based on multispectral fluorescence images. First, the influence of the maize leaf structure and the light path of the excitation light source on the fluorescence image is analysed, and the frequency domain distribution of the fluorescence information is analysed. Second, CLAHE (Contrast Limited Adaptive Histogram Equalization) and BLF-CLAHE are used to correct the fluorescence image respectively, and the images before and after correction are evaluated. Meanwhile, diagnostic models for maize chlorophyll content before and after correction were constructed. Finally, a deep learning model was established to deeply extract fluorescence information and improve the diagnostic ability of maize chlorophyll content. The study results indicate that the fluorescence yield induced by UV and red light is lower than that induced by blue light, the coefficient of variation (CV) of fluorescence intensity of maize leaves is higher in the horizontal direction than in the vertical direction, and the CV of fluorescence images of each band is higher than 0.5. Meanwhile, frequency domain analysis shows that compared with the high-frequency domain, the low-frequency domain has a higher proportion of fluorescence information, and each band is higher than 65%. Compared to traditional CLAHE, the proposed BLF-CLAHE effectively enhances the fluorescence information intensity. The chlorophyll content diagnosis model also shows that BLF-CLAHE has better chlorophyll assessment ability, the R2P of it was 0.665. The chloroph

关键词： Multispectral fluorescence imaging Unevenness correction Image enhancement Chlorophyll diagnostic model Illumination effects

来源：评论

学校读者我要写书评

暂无评论

Vaccine-YOLOv10: real-time QR code detection model for complex light condition

引用

JOURNAL OF REAL-TIME IMAGE PROCESSING 2025年第2期22卷 1-12页

作者： Zhao, Xiaobei Li, Xiang China Agr Univ Beijing Peoples R China

QR code is not only an information storage approach, but also a spatial localization sign. Compared to other spatial localization signs, QR code is more accurate and more efficient to be detected. To achieve spatial localization by QR code, detection is the essential procedure. Existing approaches perform well in regular light condition, but perform badly in complex light condition, because frame quality is extremely damaged by complex light condition. In the real world, complex light condition is very common but always unavoidable. Therefore, it is necessary and worthwhile to improve the under-complex-light QR code detection. In this paper, Vaccine-YOLOv10 (VCY) is proposed to enhance QR code detection capability in complex light condition. First, GhostConv and FasterC2f are introduced to replace the corresponding original modules of YOLOv10n. Second, Simulative Data Augment Algorithm (SDA) is proposed to simulate 5 types of complex light condition. Third, self-built Multi-Scene QR Code Dataset (MSQ) is augmented by SDA for VCY training. Compared to the baseline model YOLOv10n, VCY is improved on both lightweight and accuracy. Specifically, FPS reaches 150;GFLOPs reduces from 8.2 to 5.3;mAP50 increases from 0.877 to 0.905. Code: https://***/AlexTraveling/Vaccine-YOLOv10.

关键词： QR code detection Complex light condition YOLOv10 Data augment Spatial localization

来源：评论

学校读者我要写书评

暂无评论

An adaptive audio watermarking with frame-wise control parameter searching

引用

DIGITAL SIGNAL PROCESSING 2025年 160卷

作者： Wang, Yuzhu Xue, Yiming Liu, Xingpeng Wen, Juan China Agr Univ Beijing Peoples R China

For adaptive audio watermarking methods based on singular value decomposition (SVD), the choice of the embedding strength parameter is a crucial factor for satisfying the balance between imperceptibility and robustness. In traditional methods, the embedding strength parameter is usually the product of a constant control parameter and a variable related to the host signal. However, since the host audio signal is generally divided into frames with different energy, it is not wise to use the same control parameter to embed watermarks into different frames. In this paper, in order to achieve control parameter adaptive adjustment according to the frame energy, an adaptive audio watermarking method with frame-wise control parameter searching is developed. Firstly, the host audio signal is partitioned into paired segments through the time-sequence-segmented method. Then the watermark is embedded into the deviation value between the two largest singular values of the paired segments. We first propose an effective search algorithm called the outer loop (OL) method to find the suitable control parameter with the highest signal-to-noise ratio (SNR) under the condition of equal robustness. Then the inner- outer loop (IOL) method is further developed to find different control parameters with the lowest bit error rate (BER) while maintaining high SNR. Through the combination of the two loops, different control parameters are used to embed watermarks into different frames, contributing to an excellent balance between imperceptibility and robustness. By comparison with the existing state-of-the-art audio watermarking methods, the proposed method enhances robustness against various common attacks while ensuring high imperceptibility.

关键词： Audio watermarking Singular value decomposition Imperceptibility Robustness

来源：评论

学校读者我要写书评

暂无评论

Impact of Land Use Conversion on Soil Structure and Hydropedological Functions in an Arid Region

引用

LAND DEGRADATION & DEVELOPMENT 2025年第2期36卷 643-654页

作者： Han, Jingwen Pan, Cunzhen Sun, Yuhang Chen, Zhijun Xiong, Yunwu Huang, Guanhua China Agr Univ State Key Lab Efficient Utilizat Agr Water Resour Beijing Peoples R China China Agr Univ Chinese Israeli Int Ctr Res & Training Agr Beijing Peoples R China China Agr Univ Engn Res Ctr Agr Water Saving & Water Resources Minist Educ Beijing Peoples R China

Land use conversion critically affects soil structure and associated functions. This study investigated variations in soil structure and hydropedological characteristics across different land use types, that is, uncultivated, cultivated, and abandoned land under an arid condition. Water-stable aggregates in the uncultivated land were 15.4%-37.1% of those in the cultivated and abandoned lands at depths of 0-60cm. Reclamation of the uncultivated land enhanced soil aggregate stability, while prolonged tillage led to the loss of binding organic matter, breakdown of large aggregates and decrease in aggregate stability. The mean weight diameter of aggregates at 0-40cm depth in the cultivated land was 39.0% lower than in the abandoned land. The volume fraction of micropores (<100 mu m) in the cultivated soils was 3.7%-39.7% of that in the uncultivated soils, whereas macropores (>1000 mu m) was 1.4-1.8 times greater. Following the abandonment, soil pore diversity recovered and a hierarchical structure developed. In the abandoned land, the volume fraction of micropores (<100 mu m) was 2.4-18.9 times that of the cultivated lands, while macropores (>1000 mu m) constituted 81.4%-93.9% of those in the cultivated lands. The permeability and longitudinal dispersivity of soils in the cultivated land were significantly lower than in the uncultivated and abandoned lands, particularly at deeper soil layers. The increase in large pores in the abandoned land, important for water movement and solute transport, resulted in an order-of-magnitude rise in both permeability and longitudinal dispersivity compared with the cultivated lands. Overall, the abandoned land showed potential for rehabilitation from the perspectives of soil aggregates and pore structure. The findings may provide reference for land reclamation and management in arid regions.

关键词： land abandonment longitudinal dispersivity pore-network model soil aggregate stability soil permeability

来源：评论

学校读者我要写书评

暂无评论

Deep learning-based model to classify mastitis in Holstein dairy cows

引用

BIOSYSTEMS ENGINEERING 2025年 252卷 92-104页

作者： Chu, Mengyuan Si, Yongsheng Li, Qian Liu, Xiaowen Liu, Gang China Agr Univ Key Lab Smart Agr Syst Minist Educ Beijing 100083 Peoples R China China Agr Univ Key Lab Agr Informat Acquisit Technol Minist Agr & Rural Affairs China Beijing 100083 Peoples R China Hebei Agr Univ Coll Informat Sci & Technol Baoding Hebei Peoples R China

The occurrence and prevalence of dairy cow mastitis has brought significant challenges to animal welfare and economy. To overcome the complexities and accumulated errors present in previous detection methods, a rapid and accurate mastitis detection approach is developed based on image processing and deep learning, leveraging thermal infrared imaging. Image processing techniques, including the Hough transform and morphological operations, are used to classify affected cows from thermal images. An image pyramid is constructed based on upsampling to tackle motion blur induced by the cows' rapid movement. The multi-scale convolution and the spatial and channel Squeeze & Excitation (scSE) block were integrated into the DenseNet-201 architecture to enhance the feature extraction process. This enabled the network to adaptively recalibrate channel-wise feature responses and strengthening the discriminative power of the learned representations. For mastitis detection, a deep learning model, the multi-scale scSE-DenseNet-201 (MS-scSE-DenseNet-201) architecture, is refined to predict the severity of mastitis. The framework takes images of both sides of the cow's udder as input, and outputs one of three mastitis severity levels: negative (N), subclinical mastitis (SCM), or clinical mastitis (CM). To assess the model's performance in detecting mastitis, a dataset comprising 5000 thermal images from 802 cows, was used. The model achieved accuracy, precision, and recall of 90.18%, 92.16%, and 88.38%, respectively, showing notable improvement over previous methods. This work integrated object segmentation and blind deblurring to strengthen the MS-scSE-DenseNet-201 in the automatic detection of cow mastitis, which will open a promising application horizon for other animal disease diagnostics.

关键词： Livestock Computer vision Image processing Object segmentation Motion deblurring

来源：评论

学校读者我要写书评

暂无评论

Cascaded Gravity-Driven Microfluidic Chip with Tilt-Actuated Siphon Valves for Rapid Detection of Salmonella Typhimurium

引用

ANALYTICAL CHEMISTRY 2025年第13期97卷 7231-7241页

作者： Yang, Fengzhen Wang, Tianchang Wang, Siyuan Li, Meixuan Jia, Kaiyuan Ding, Ying Lin, Jianhan Xi, Xinge China Agr Univ Key Lab Agr Informat Acquisit Technol Minist Agr & Rural Affairs Beijing 100083 Peoples R China China Agr Univ Key Lab Smart Agr Syst Integrat Minist Educ Beijing 100083 Peoples R China

In recent years, urgent food safety issues have heightened the demand for rapid detection technologies for foodborne pathogens, especially biosensors featuring simplicity, rapidity, and high sensitivity. Yet despite a booming surge in related published studies, commercializing these biosensors remains a constant challenge and persistent objective for researchers. In this study, a gravity-driven microfluidic chip with tilt-actuated siphon valves was developed, integrating silica magnetic beads based nucleic acid separation and recombinase-aided amplification (RAA) detection of Salmonella Typhimurium by simple operations along with a portable biosensing device. By chip inclination, the balance between gravity of reagents and capillary force around siphon valves is altered;thus, simple tilting operations could provide a driving force for fluid flow without the need for external pumps. Siphon valves were designed geometrically with surface modifications to generate comfortable tilting angles, and a chip holder with a built-in angle guide was designed to ensure consistent operational performance. A smartphone app was developed to monitor fluorescence signals for quantitative detection of S. Typhimurium. Experimental results showed that this portable biosensing device could detect S. Typhimurium in spiked chicken samples at concentrations as low as 1.1 x 101 CFU/mL within 60 min, with recovery rates ranging from 91.54% to 117.27%. The siphon valves also demonstrated compatibility with diverse liquid properties, offering scalability and adaptability for various detection scenarios and potential for the detection of various pathogens in food safety and clinical diagnostics by using related nucleic acid probes and reagents.

关键词： Salmonella

来源：评论

学校读者我要写书评

暂无评论

Effect of sowing date on physicochemical properties of waxy and non-waxy proso millet (Panicum miliaceum L.) starches

引用

INTERNATIONAL JOURNAL OF BIOLOGICAL MACROMOLECULES 2025年 303卷 140626页

作者： Ren, Jiangling Liu, Yuhan Liu, Sichen Wang, Ruiyun Qiao, Zhijun Cao, Xiaoning Shanxi Agr Univ Ctr Agr Genet Resources Res Taiyuan 030031 Peoples R China Shanxi Agr Univ Coll Agr Taigu 030801 Peoples R China

In this experiment, waxy (Shuzi, P1) and non-waxy (Ningmei No. 14, P2) proso millet were used as experimental materials to study the quality changes of waxy and non-waxy proso millet in different sowing dates as well as the differences in the physical and chemical properties of starch, such as starch morphological structure, crystal structure and gelatinization properties. The results showed that with the postponement of the sowing date, the total starch contents of P1 and P2 decreased by 2 % - 7.28 % and 3.26 % - 8.23 %, respectively, and the protein contents decreased by 0.1 % - 9.91 % and 2.52 % - 5.03 %, respectively. Compared with B1, B3 - B5 reduced the amylose content of P1 by 15.21 % - 26.80 %. With the postponement of the sowing date, the breakdown (BD) of P1 increased by 4.68-22.79 %, while the trough viscosity (TV) and final viscosity (FV) decreased by 2.50 % 17.43 % and 2.58 % - 9.21 %, respectively. The peak viscosity (PV), TV, BD and FV of P2 increased by 10.33 % 36.95 %, 9.31 % - 19.86 %, 12.68 % - 81.07 % and 5.69 % - 36.46 %, respectively, with the postponement of the sowing date. The sowing date also affected the volume distribution of proso millet grains. This study clarified the influence of sowing date on the quality of proso millet grains and the physical and chemical properties of starch, providing a theoretical basis for improving the high-yield and high-quality cultivation techniques of proso millet grains and the deep processing of products.

关键词： Proso millet Sowing date Physicochemical properties of starch Quality

来源：评论

学校读者我要写书评

暂无评论

Predicting the nonlinear dynamics of spatiotemporal PDEs via physics-informed informer networks

引用

NONLINEAR DYNAMICS 2025年第10期113卷 10929-10946页

作者： Zhang, Ziyang Zhang, Feifan Chen, Tailai Zhang, Jiachen China Agr Univ Coll Sci Beijing Peoples R China China Agr Univ Coll Informat & Elect Engn Beijing Peoples R China

Partial differential equations (PDEs) have been widely used in physics, engineering, finance, and other fields to simulate various real-world phenomena. Recent advances in deep learning have shown the great potential of physics-informed neural networks as a novel machine learning model that combines deep learning with the laws of physics. However, most of the existing PINNs methods based on fully connected neural networks only focus on the spatial connection of the loss function of low-dimensional space-time, which constitutes an inherent limitation. To this end, coupled neural networks of Informer and PINNs called PhyInformer are proposed, which take into account the dependencies of data on time and space. The loss function is defined as the residuals of the discretized PDEs together with its boundary value condition loss and initial value condition loss. Extensive numerical experiments on four different models (e.g., 2D-Burgers' equations, 2D-Diffusion equations and Allen-Cahn equations and Wave equations). Demonstrate that our proposed method significantly outperforms existing PINNs.

关键词： PINNs Deep learning Informer

来源：评论

学校读者我要写书评

暂无评论

Large Language Model Based Intelligent Fault Information Retrieval System for New Energy Vehicles

引用

APPLIED SCIENCES-BASEL 2025年第7期15卷 4034-4034页

作者： Zhang, Haiyu Zhao, Yinghui Sun, Boyu Wu, Yaqi Fu, Zetian Xiao, Xinqing China Agr Univ Yantai Inst Yantai 264670 Peoples R China China Agr Univ Coll Engn Beijing 100083 Peoples R China

In recent years, the rapid development of the new energy vehicle (NEV) industry has exposed significant deficiencies in intelligent fault diagnosis and information retrieval technologies, especially in intelligent fault information retrieval, which faces persistent challenges including inadequate system adaptability and reasoning bottlenecks. To address these challenges, this study proposes a Retrieval-Augmented Generation (RAG) framework that integrates large language models (LLMs) with knowledge graphs (KGs). The framework consists of three key components: fault data collection, knowledge graph construction, and fault knowledge model training. The primary research contributions are threefold: (1) A domain-optimized fine-tuning strategy for LLMs based on NEV fault characteristics, verifying the superior accuracy of the Bidirectional Encoder Representations from Transformers (BERT) model in fault classification tasks. (2) A structured knowledge graph encompassing 122 fault categories, developed through the ChatGLM3-6B model completing named entity and knowledge relation extraction to generate fault knowledge and build a paraphrased vocabulary. (3) An intelligent fault information retrieval system that significantly outperforms traditional models in NEV-specific Q&A scenarios, providing multi-level fault cause analysis and actionable solution recommendations.

关键词： new energy vehicle fault intelligent retrieval system generative language models knowledge graphs retrieval-augmented generation

来源：评论

学校读者我要写书评

暂无评论

Ground straw mulching level classification based on a terral grid system and deep learning

引用

COMPUTERS AND ELECTRONICS IN agrICULTURE 2025年 230卷

作者： Jiang, Shan Li, Hongwen Zhang, Zhao Liu, Kaidong Lu, Caiyun Wang, Chao Saha, Chayan Kumer Li, Rongrong Wu, Zhengyang Yang, Zongfu He, Dong China Agr Univ Coll Engn Beijing 100083 Peoples R China China Agr Univ State Key Lab Efficient Utilizat Agr Water Resourc Beijing 100083 Peoples R China China Agr Univ Key Lab Smart Agr Syst Integrat Minist Educ Beijing 100083 Peoples R China China Agr Univ Key Lab Agr Informat Acquisit Technol Minist Agr & Rural Affairs Beijing 100083 Peoples R China State Key Lab Intelligent Agr Power Equipment Beijing 100083 Peoples R China Bangladesh Agr Univ Dept Farm Power & Machinery Mymensingh 2202 Bangladesh

The detection of ground straw mulching levels plays a crucial role in implementing conservation tillage efficiently. To address the need for quick and accurate determination of straw mulching levels, a rapid detection method based on deep learning was proposed, which consists of a terral grid system, semi-automatic straw mulching grading based on ASPP-CBAM MobileNet (AC-MobileNet), and a deep residual network ResNet101. The terral grid system was applied to obtain local sliced images at the grid intersections from straw mulching images. Then, the AC-MobileNet model was used to acquire the type of local sliced images automatically, making it obtain the true value of each straw mulching level and establish the dataset quickly. Subsequently, the deep residual network ResNet101 was utilized for straw mulching level detection. The test results showed that the ACMobileNet model achieved a sliced image classification accuracy of 96.3%, surpassing the MobileNetv3 network model by 2.1%. In comparison with classification networks such as AlexNet, ShuffleNetv1, and EfficientNetv2, the AC-MobileNet model exhibited the highest accuracy. The accuracy of ResNet101 in straw mulching level classification was 98.3%, outperforming DenseNet161, EfficientNetv2, and VGG16 networks. The proposed method demonstrated a detection speed of 750 times higher than that of manual detection in the field, keeping the straw mulching level classification results aligned with the manual results. The developed methodology paves the road for accurate and rapid detection of ground straw mulching levels.

关键词： Image processing AC-MobileNet ResNet101 Straw mulching level Terral grid system

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：