检索结果-内蒙古大学图书馆

Quantifying consistency of crop establishment using a lightweight U-Net deep learning architecture and image processing techniques

引用

COMPUTERS AND ELECTRONICS IN AGRICULTURE 2024年 217卷

作者： Ullah, Muhib Islam, Fatimah Bais, Abdul Univ Regina Fac Engn & Appl Sci 3737 Wascana Pkwy Regina SK S4S 0A2 Canada

Consistency of crop establishment is a measure of uniformity of crop attributes, such as plant stand count, crop emergence rate, and plant spacing across the field. Quantifying consistency during the early crop growth stage is important for establishment decisions to use targeted nutrients and to facilitate timely replanting in inconsistent crop regions. Crop consistency can be analysed using two key parameters: plant stand count and spacing statistics since they provide insight into plant density and its emergence percentage. However, manual assessment of them is time-consuming, prone to errors, and labour-intensive in large fields. An alternative method is proposed to automate estimating these parameters using field imagery under uncontrolled settings. We use the YOLOv5-based object detection model for plant counting, which attains a mean average precision of 0.956 to detect Canola plants. A Lightweight U -Net model is proposed to segment rows, followed by Guo-Hall thinning and Probabilistic Hough Transform to determine inter -row and inter -plant spacing. Our proposed row segmentation model achieves a mean Intersection over Union (mIoU) of 0.8444 with class -wise IoU of 0.9925 and 0.6963 for background and crop using fewer parameters. The new architecture uses only 14M parameters and achieves performance comparable to the state-of-the-art U -Net (32.5M) and SegNet (29M).

关键词： Lightweight U-net Crop consistency Probabilistic hough transform Guo-Hall thinning Plant density estimation

来源：评论

学校读者我要写书评

暂无评论

image processing and deep learning Based Road Object Detection System for Safe Transportation 15

Image Processing and Deep Learning Based Road Object Detecti...

引用

15th International Conference on Computing Communication and Networking Technologies, ICCCNT 2024

作者： Prito, Rizvee Hassan Hossain, Md. Shafayat Islam, Md. Shajibul Meem, Mehzabin Ali, Md. Nawab Yousuf East West University Computer Science and Engineering Dept Dhaka Bangladesh

ISBN: (纸本)9798350370249

Road object detection, a pivotal task in computer vision and artificial intelligence, is dedicated to the identification and precise localization of a diverse array of elements on roadways, including vehicles, pedestrians, road signs, traffic lights, and potential obstacles. The essence of this task lies in its ability to provide real-time and precise object detection, ultimately serving as a crucial safeguard to prevent accidents and ensure the safety of drivers, passengers, and pedestrians. It also lays the foundation for advanced warning systems and aids in collision avoidance. Several popular models were implemented, encompassing YOLOv7, YOLOv7-Modified, YOLOv7-Tiny, YOLOv7-E6E, Faster R-CNN, and SSD. Among these, YOLOv7 achieved an impressive mean average precision (mAP) of 83.6%, with an inference speed of 15.1 ms, while YOLOv7E6E achieved the highest mAP of 86.2%, but with the cost of a slower inference speed of 30.3 ms. The modified version of YOLOv7 has produced 2.2% higher average precision accuracy (89.1%) than the main version of YOLOv7 due to its double RepVGG layers, skip connection and concatenation layers in the head architecture. To make this research accessible to a wider audience, a user-friendly web application is developed with an intuitive interface. © 2024 IEEE.

关键词： Traffic signs

来源：评论

学校读者我要写书评

暂无评论

Supervised deep learning based Leaf Disease and Pest Detection using image processing 7

Supervised Deep Learning based Leaf Disease and Pest Detecti...

引用

7th International Conference on Intelligent Computing and Control Systems, ICICCS 2023

作者： Isravel, Deva Priya Somasundaram, Kumaresan Jestin Josephraj, M. Christopher Paul, L. Johnson, J.P. Karunya Institute of Technology and Sciences Department of Computer Science and Engineering Tamil Nadu Coimbatore India Karunya Institute of Technology and Sciences Department of Computer Science and Engineering Tamil Nadu Sivaganga India Karunya Institute of Technology and Sciences Department of Computer Science and Engineering Tamil Nadu Kumbakonam India Karunya Institute of Technology and Sciences Department of Computer Science and Engineering Tamil Nadu Dharmapuri India

ISBN: (纸本)9798350397253

The key problems that influence plant health and crop yield quality are leaf disease and pests. To improve crop production many advanced technologies are deployed. One of the predominant technologies is the incorporation of deep learning (DL) techniques. DL supports numerous training and testing models that are suitable for smart agriculture. real-time detection of plant diseases and pest detection is made simpler and more efficient by using DL models. This paper provides an extensive survey of the DL techniques associated with agriculture automation and discusses the latest models focusing on leaf disease detection and pest identification. To select the best features and to optimize the accuracy of the results, an optimizer is identified and an enhanced deep learning model is proposed. The VGG16 and YOLOV5s models are deployed with ADAM optimizer. The results illustrate that the proposed optimized approach achieves an accuracy of 98.71% for leaf disease detection and 97.52% for pest detection. © 2023 IEEE.

关键词： image processing

来源：评论

学校读者我要写书评

暂无评论

Pioneering Early Parkinsons Disease Detection with deep learning and image processing Innovations 10

Pioneering Early Parkinsons Disease Detection with Deep Lear...

引用

10th International Conference on Advanced Computing and Communication Systems, ICACCS 2024

作者： Srinivasarao, Narisetty Khan, Mohammad Basith Anwar Haneesha, Pullipati Meghana, Palaparthi Department of Computer Science and Engineering Mylavaram India

ISBN: (纸本)9798350384369

Parkinson's disease, a neurological disorder which affects the nervous system, manifests as unintentional and uncontrollable movements in the body. With over 6 million individuals globally affected, early detection of PD proves challenging, given its varied symptoms. The manifestations of the disease can differ among individuals, with symptoms typically initiating as a tremor in one hand and progressively impacting the entire body over time. Typically, symptoms begin with a hand tremor and progress throughout the body. Currently, conventional clinical approaches do not possess the capability to detect Parkinson's disease during its initial phases. Diagnosis relies on medical history, brain MRI images, and manual symptom observation, a process that is time-consuming and often insufficient for early detection. The potential of deep learning for early PD detection is explored using datasets comprising spiral drawings from both normal and PD-affected individuals. CNN and ResNet50 algorithms are applied to spiral drawings, enhancing the model's ability to capture intricate patterns. The training process utilizes 60% of the dataset, with the remaining 40% reserved for testing. In the case of spiral drawings, the ResNet50 algorithm proves instrumental. It transforms spiral drawings into pixels, facilitating classification. During testing, the pixels of current drawings are compared with previously trained CNN and ResNet50 models to identify the disease. The amalgamation of results from ResNet50-enhances spiral drawing datasets significantly and improves disease detection accuracy more than CNN. Individual's data can be given as input into the system for efficient disease detection. © 2024 IEEE.

关键词： Ability testing

来源：评论

学校读者我要写书评

暂无评论

Lane detection using the concept of deep learning and Digital image processing

Lane detection using the concept of Deep learning and Digita...

引用

Electrical Electronics and Computing Technologies (ICEECT), International Conference on

作者： Shashi Kant Ojha Abhishek Kumar Sandeep Kumar CSE Department Sharda University Greater Noida

ISBN: (数字)9798350378092

ISBN: (纸本)9798350378108

Lane detection technology plays a pivotal role in enabling autonomous navigation in vehicles. However, existing systems primarily cater to well-structured roads with clear lane markings, rendering them ineffective in scenarios where markings are unclear or absent. This study critically evaluates an existing approach for detecting lanes on unmarked roads, followed by the proposal of an enhanced methodology. Both approaches leverage digital image processing techniques and rely solely on vision or camera data. The primary objective is to derive real-time curvature values to facilitate driver-assistance systems in making necessary turns and preventing vehicles from veering off-road.

关键词： deep learning Computers Lane detection Navigation Roads Digital images Rendering (computer graphics) Cameras real-time systems Proposals

来源：评论

学校读者我要写书评

暂无评论

deep learning-based image processing framework for efficient surface litter detection in Computer Vision applications

引用

JOURNAL OF RADIATION RESEARCH AND APPLIED SCIENCES 2025年第2期18卷

作者： Shad, Ifza Zhang, Zuping Asim, Muhammad Al-Habib, Mohammed Chelloug, Samia Allaoua Abd El-Latif, Ahmed Cent South Univ Sch Comp Sci & Engn Changsha 410083 Peoples R China Prince Sultan Univ Coll Comp & Informat Sci EIAS Data Sci Lab Riyadh 11586 Saudi Arabia Prince Sultan Univ Ctr Excellence Quantum & Intelligent Comp Riyadh 11586 Saudi Arabia Princess Nourah bint Abdulrahman Univ Coll Comp & Informat Sci Dept Informat Technol POB 84428 Riyadh 11671 Saudi Arabia Menoufia Univ Fac Sci Dept Math & Comp Sci Shibin Al Kawm 32511 Egypt

Computer vision plays a vital role in automating environmental analysis by enabling real-time object detection and classification in diverse conditions. Litter pollution poses significant health and environmental risks due to inefficient disposal, and manual oversight is labor intensive. Effective litter detection is crucial for large-scale environmental monitoring. However, existing models face challenges such as the complexity of detecting shadowy objects in varying lighting conditions (e.g., during rain or under sun rays), difficulty in recognizing small objects, low accuracy, and poor real-time performance. Existing two-stage detectors, such as Faster R-CNN, also struggle with these issues. This paper introduces an automated deep learning-based image processing approach for accurate litter detection across different locations, using an enhanced version of YOLOv9s called LD-YOLOv9s. Key improvements in this novel approach include replacing convolutional layers with DynConvLayer in the backbone, integrating an SDConv-ADown module to substitute down-sampling layers in the neck, and using MPD-IoU instead of CIoU. These modifications reduce the chances of overlooking small objects, such as caps or lids, which had the least class meaning in the dataset, achieving a mAP of 78.3% with an inference time of 6.7ms. A significant contribution of this work is the LD-2024 dataset, curated from indoor and outdoor environments with manually annotated images. Performance comparisons were made with several YOLO versions (YOLOv3, YOLOv5, YOLOv6, YOLOv7, YOLOv8, YOLOv9 variants) and traditional object detectors (Faster R-CNN, Center Net, Retina Net, Cascade R-CNN). Ablation studies validate the effectiveness of LD-YOLOv9s, which outperforms conventional methods, achieving a 6.3% improvement in mean average precision (mAP) over YOLOv9s on the LD-2024 dataset.

关键词： deep learning Computer vision image processing Litter Detection YOLO DynConvLayer SDConv MPD-IoU LD-YOLOv9s Object Detection

来源：评论

学校读者我要写书评

暂无评论

A Novel deep learning and image processing Based Technique for Lumpy Skin Disease Detection 2

A Novel Deep Learning and Image Processing Based Technique f...

引用

2nd International Conference on Advancements in Smart, Secure and Intelligent Computing, ASSIC 2024

作者： Singh, Jay Prakash Ghosh, Debolina Singh, Jagannath Manipal University Jaipur Department of Comp. Sci. and Engg. Jaipur India Manipal University Jaipur Department of Information Technology Jaipur India School of Comp.Engg. Kiit Deemed to Be University Bhubaneswar India

ISBN: (纸本)9798350370188

Lumpy skin disease is a very infectious disease in cattle. Every year many cattle die because of lumpy skin disease. Recently, Indian states especially Rajasthan, Haryana, Punjab have witnessed the death of many cows due to lumpy skin infection. This disease becomes very dangerous and spreads at a high rate if not treated well at the time of infection. A proper technique is required to detect the lumpy skin infection in cattle so that proper diagnosis can be adopted to stop the infection from spreading and to save the life of the infected cattle. With the emergence of machine learning and deep learning approaches it seems that detection can be done at the inception of infection in the cattle. Many researchers have suggested various ways to detect the lumpy skin disease but there is need of an approach which can guarantee the high accuracy, precision, recall and f1 score for the lumpy skin infection. We have used some of the advanced pre-trained deep learning models and also proposed a modified convolution neural network model for lumpy skin infection detection. We collected dataset from the online platform. We collected total 936 cattle images including healthy and lumpy skin infected cattle. Out of 936 cattle images 421 were lumpy affected and rest 515 were images of the healthy cattle. We split the image in training and testing dataset. We used 80 percent of images for training the models and 20 % of images for testing the performance of the models. We trained InceptionV3, VGG16, ResNet50V2 and modified CNN model under similar condition by properly tuning the hyper parameters. We have observed that proposed modified CNN model outperformed all the reset three models. On training data, modified CNN yield an accuracy of 99.01 %, InceptionV3 furnished an accuracy of 98.85 %, Resnet50V2 has an accuracy of 92.09% and VGG16 guaranteed an accuracy of 92%. On testing data, proposed modified CNN, InceptionV3, ResNet50V2 and VGG 16 has accuracy 97.87 %, 97.34%, 93.09% and 93

关键词： Statistical tests

来源：评论

学校读者我要写书评

暂无评论

deep learning models for digital image processing: a review

引用

ARTIFICIAL INTELLIGENCE REVIEW 2024年第1期57卷 11-11页

作者： Archana, R. Jeevaraj, P. S. Eliahim Bharathidasan Univ Bishop Heber Coll Dept Comp Sci Tiruchirappalli 620017 Tamil Nadu India

Within the domain of image processing, a wide array of methodologies is dedicated to tasks including denoising, enhancement, segmentation, feature extraction, and classification. These techniques collectively address the challenges and opportunities posed by different aspects of image analysis and manipulation, enabling applications across various fields. Each of these methodologies contributes to refining our understanding of images, extracting essential information, and making informed decisions based on visual data. Traditional image processing methods and deep learning (DL) models represent two distinct approaches to tackling image analysis tasks. Traditional methods often rely on handcrafted algorithms and heuristics, involving a series of predefined steps to process images. DL models learn feature representations directly from data, allowing them to automatically extract intricate features that traditional methods might miss. In denoising, techniques like Self2Self NN, Denoising CNNs, DFT-Net, and MPR-CNN stand out, offering reduced noise while grappling with challenges of data augmentation and parameter tuning. image enhancement, facilitated by approaches such as R2R and LE-net, showcases potential for refining visual quality, though complexities in real-world scenes and authenticity persist. Segmentation techniques, including PSPNet and Mask-RCNN, exhibit precision in object isolation, while handling complexities like overlapping objects and robustness concerns. For feature extraction, methods like CNN and HLF-DIP showcase the role of automated recognition in uncovering image attributes, with trade-offs in interpretability and complexity. Classification techniques span from Residual Networks to CNN-LSTM, spotlighting their potential in precise categorization despite challenges in computational demands and interpretability. This review offers a comprehensive understanding of the strengths and limitations across methodologies, paving the way for informed decis

关键词： image processing deep learning models Convolutional neural networks (CNN)

来源：评论

学校读者我要写书评

暂无评论

Developing Forest Road Recognition Technology Using deep learning-Based image processing

引用

FORESTS 2024年第8期15卷 1469页

作者： Lee, Hyeon-Seung Kim, Gyun-Hyung Ju, Hong Sik Mun, Ho-Seong Oh, Jae-Heun Shin, Beom-Soo Natl Inst Forest Sci Forest Technol & Management Res Ctr Pochon 11187 South Korea Kangwon Natl Univ Dept Biosyst Engn 1 Kangwondaehak Gil Chunchon 24341 South Korea Kangwon Natl Univ Grad Sch Interdisciplinary Program Smart Agr 1 Kangwondaehak Gil Chunchon 24341 South Korea

This study develops forest road recognition technology using deep learning-based image processing to support the advancement of autonomous driving technology for forestry machinery. images were collected while driving a tracked forwarder along approximately 1.2 km of forest roads. A total of 633 images were acquired, with 533 used for the training and validation sets, and the remaining 100 for the test set. The YOLOv8 segmentation technique was employed as the deep learning model, leveraging transfer learning to reduce training time and improve model performance. The evaluation demonstrates strong model performance with a precision of 0.966, a recall of 0.917, an F1 score of 0.941, and a mean average precision (mAP) of 0.963. Additionally, an image-based algorithm is developed to extract the center from the forest road areas detected by YOLOv8 segmentation. This algorithm detects the coordinates of the road edges through RGB filtering, grayscale conversion, binarization, and histogram analysis, subsequently calculating the center of the road from these coordinates. This study demonstrates the feasibility of autonomous forestry machines and emphasizes the critical need to develop forest road recognition technology that functions in diverse environments. The results can serve as important foundational data for the future development of image processing-based autonomous forestry machines.

关键词： autonomous forestry machines image processing deep learning YOLO

来源：评论

学校读者我要写书评

暂无评论

Enhancing Palm Recognition Accuracy Through image processing deep learning Techniques

Enhancing Palm Recognition Accuracy Through Image Processing...

引用

Data Science and Network Security (ICDSNS), International Conference on

作者： J. Sheela Mercy S. Silvia Priscila Department of Computer Applications Bharath Institute of Higher Education and Research Chennai India Department of Computer Science Bharath Institute of Higher Education and Research Chennai India

ISBN: (数字)9798350373110

ISBN: (纸本)9798350373127

Palm recognition systems play an important role in biometric authentication; however, existing systems frequently have low accuracy and resiliency due to problems such as changing lighting conditions, occlusions, and hand orientations. The paper describes a novel way to improve palm recognition accuracy by combining modern image processing and deep learning approaches. Unlike traditional systems that utilize handcrafted features and shallow learning algorithms, the proposed system uses convolutional neural networks (CNNs) for feature extraction and classification, as well as advanced image preprocessing techniques. The technology preprocesses palm images to increase quality and consistency, thereby reducing the impact of environmental factors. Cross-validation trials yielded consistent results, with an average accuracy of 95%, precision of 92%, recall of 96%, and Fl-score of 94%. Furthermore, computational efficiency comparisons reveal that the proposed system beats existing ones in terms of training time, inference time, and model size, indicating that it is a promising option for accurate and efficient palm recognition in a variety of real-world applications.

关键词： Training deep learning Accuracy image recognition Computational modeling Lighting Biometric authentication

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：