检索结果-内蒙古大学图书馆

FPGA-Based Real-Time object detection and classification System Using YOLO for Edge Computing

IEEE ACCESS 2024年 12卷 73268-73278页

作者： Al Amin, Rashed Hasan, Mehrab Wiese, Veit Obermaisser, Roman Univ Siegen Inst Embedded Syst D-57076 Siegen Germany

The leap forward in research progress in real-time object detection and classification has been dramatically boosted by including Embedded Artificial Intelligence (EAI) and Deep Learning (DL). Real-time object detection and classification with deep learning require many resources and computational power, which makes it more difficult to use deep learning methods on edge devices. This paper proposed a new, highly efficient Field Programmable Gate Array (FPGA) based real-time object detection and classification system using You Only Look Once (YOLO) v3 Tiny for edge computing. However, the proposed system has been instantiated with Advanced Driving Assistance Systems (ADAS) for evaluation. Traffic light detection and classification are crucial in ADAS to ensure drivers' safety. The proposed system used a camera connected to the Kria KV260 FPGA development board to detect and classify the traffic light. Bosch Small Traffic Light Dataset (BSTLD) has been used to train the YOLO model, and Xilinx Vitis AI has been used to quantify and compile the YOLO model. The proposed system can detect and classify traffic light signals from a high-definition (HD) video streaming in 15 frames per second (FPS) with 99% accuracy. In addition, it consumes only 3.5W power, demonstrating the ability to work on edge devices. The on-road experimental results represent fast, precise, and reliable detection and classification of traffic lights in the proposed system. Overall, this paper demonstrates a low-cost and highly efficient FPGA-based system for real-time object detection and classification.

关键词： FPGAs object detection and classification YOLO edge computing

来源：评论

学校读者我要写书评

暂无评论

Robotic Guide Dog for Real-time Indoor object detection and classification with Localization

Robotic Guide Dog for Real-time Indoor Object Detection and ...

引用

25th IEEE Applied Sensing Conference (IEEE APSCON)

作者： Rees, Nathan Thiyagarajan, Karthick Kodagoda, Sarath Univ Technol Sydney UTS Robot Inst Sydney NSW Australia

ISBN: (纸本)9798350317275

Guide dog robots with advanced sensing abilities could be a big boon to vision-impaired people as some of them may choose technological solutions over real-life guide dogs. In this study, we propose a method that combines a robotic guide dog sensing system with the YOLO-GUIDE framework to enable real-time indoor object detection and classification with localization. The performance was assessed using ten indoor objects. The qualitative test outcomes showed the effectiveness of the proposed method, while quantitative evaluation results with 0.76 Precision, 0.67 Recall, and a 0.71 F1-score indicate high performance. The YOLO-GUIDE proved its superiority by outperforming other relevant models.

关键词： Sensor applications blind assistive robotics YOLO object detection and classification object localization

来源：评论

学校读者我要写书评

暂无评论

Nature-Inspired Search Method and Custom Waste object detection and classification Model for Smart Waste Bin

引用

SENSORS 2022年第16期22卷 6176页

作者： Agbehadji, Israel Edem Abayomi, Abdultaofeek Bui, Khac-Hoai Nam Millham, Richard C. Freeman, Emmanuel Durban Univ Technol Fac Accounting & Informat POB 1334 ZA-4000 Durban South Africa Mangosuthu Univ Technol Dept Informat & Commun Technol POB 12363 ZA-4026 Durban South Africa Korea Inst Sci & Technol Informat KISTI Supercomp Applicat Ctr Daejeon 34141 South Korea Durban Univ Technol Dept Informat Technol ICT & Soc Res Grp POB 1334 ZA-4000 Durban South Africa Ghana Commun Technol Univ Dept Comp Sci PMB 100 Accra Ghana

Waste management is one of the challenges facing countries globally, leading to the need for innovative ways to design and operationalize smart waste bins for effective waste collection and management. The inability of extant waste bins to facilitate sorting of solid waste at the point of collection and the attendant impact on waste management process is the motivation for this study. The South African University of Technology (SAUoT) is used as a case study because solid waste management is an aspect where SAUoT is exerting an impact by leveraging emerging technologies. In this article, a convolutional neural network (CNN) based model called You-Only-Look-Once (YOLO) is employed as the object detection algorithm to facilitate the classification of waste according to various categories at the point of waste collection. Additionally, a nature-inspired search method is used as learning rate for the CNN model. The custom YOLO model was developed for waste object detection, trained with different weights and backbones, namely ***.74, darknet19_***.23, ***.137 and ***.29, respectively, for Yolov3, Yolov3-tiny, Yolov4 and Yolov4-tiny models. Eight (8) classes of waste and a total of 3171 waste images are used. The performance of YOLO models is considered in terms of accuracy of prediction (Average Precision-AP) and speed of prediction measured in milliseconds. A lower loss value out of a percentage shows a higher performance of prediction and a lower value on speed of prediction. The results of the experiment show that Yolov3 has better accuracy of prediction as compared with Yolov3-tiny, Yolov4 and Yolov4-tiny. Although the Yolov3-tiny is quick at predicting waste objects, the accuracy of its prediction is limited. The mean AP (%) for each trained version of YOLO models is Yolov3 (80%), Yolov4-tiny (74%), Yolov3-tiny (57%) and Yolov4 (41%). This result of mAP (%) indicates that the Yolov3 model produces the best performance results (8

关键词： smart bin Internet of Things (IoT) enabled object detection and classification convolutional neural network (CNN) You-Only-Look-Once (YOLO) Kestrel-based search algorithm (KSA)

来源：评论

学校读者我要写书评

暂无评论

Comparing deep learning models for low-light natural scene image enhancement and their impact on object detection and classification: Overview, empirical evaluation, and challenges

引用

SIGNAL PROCESSING-IMAGE COMMUNICATION 2022年第0期109卷

作者： Al Sobbahi, Rayan Tekli, Joe Lebanese Amer Univ LAU Dept Elect & Comp Engn 36 Byblos Mt Lebanon Lebanon Univ Pay & Pays Adour UPPA LIUPPA Lab SPIDER Res Team F-64600 Anglet Aquitaine France

Low-light image (LLI) enhancement is an important image processing task that aims at improving the illumination of images taken under low-light conditions. Recently, a remarkable progress has been made in utilizing deep learning (DL) approaches for LLI enhancement. This paper provides a concise and comprehensive review and comparative study of the most recent DL models used for LLI enhancement. To our knowledge, this is the first comparative study dedicated to DL-based models for LLI enhancement. We address LLI enhancement in two ways: (i) standalone, as a separate task, and (ii) end-to-end, as a pre-processing stage embedded within another high-level computer vision task, namely object detection and classification. The paper consists of six logical parts. First, we provide an overview of the background and literature in LLI enhancement. Second, we describe the test data and experimental setup of the study. Third, we present a quantitative and qualitative comparison of the visual and perceptual quality achieved by 10 of the most recent DL-based LLI enhancement models. Fourth, we present a comparative analysis for object detection and classification performance achieved by 4 different object detection models applied on LLIs and their enhanced counterparts. Fifth, we perform a feature analysis of DL feature maps extracted from normal, low-light, and enhanced images, and perform the occlusion experiment to better understand the effect of LLI enhancement on the object detection and classification task. Finally, we provide our conclusions and highlight future steps and potential directions.

关键词： Image enhancement Low-light conditions Deep learning models object detection and classification Empirical comparison

来源：评论

学校读者我要写书评

暂无评论

A Miniaturized and Intelligent Lensless Holographic Imaging System With Auto-Focusing and Deep Learning-Based object detection for Label-Free Cell classification

IEEE PHOTONICS JOURNAL

引用

IEEE PHOTONICS JOURNAL 2024年第3期16卷 1页

作者： Chen, Jin Han, Wentao Fu, Liangzun Lv, Zhihang Chen, Haotian Fang, Wenjing Hou, Jiale Yu, Haohan Huang, Xiwei Sun, Lingling Hangzhou Dianzi Univ Key Lab RF Circuits & Syst Minist Educ Hangzhou 310018 Peoples R China

Cell detection and classification is a key technique for disease diagnosis, but conventional methods such as optical microscopy and flow cytometry have limitations in terms of field-of-view (FOV), throughput, cost, size, and operation complexity. Lensless holographic imaging is a promising alternative that offers large FOV, rich information content, and simple structure. However, its performance on cell detection and classification still needs to be improved. In this paper, we propose an intelligent cell detection system based on lensless holographic imaging and deep learning. Our system uses unstained cells suspended in solution as samples and employs a threshold segmentation-based auto-focusing algorithm to determine the optimal focusing distance for each imaging session. We also use a deep learning-based object detection neural network to classify different types of cells from the focused holographic images without the need for cell segmentation. We demonstrated the performance of our system using four cell detection tasks: tumor cells vs. polystyrene microspheres (77.6% accuracy), different tumor cells (80.1% accuracy), red blood cells vs. white blood cells (78.1% accuracy), white blood cell subtypes (88% accuracy), which showed that our system achieved high accuracy with label-free, portable, intelligent, and fast cell detection capabilities. It has potential applications in the miniaturized cell detection field.

关键词： Auto-focusing algorithm deep learning lensless holographic imaging object detection and classification label-free

来源：评论

学校读者我要写书评

暂无评论

Seafloor debris detection using underwater images and deep learning-driven image restoration: A case study from Koh Tao, Thailand

引用

MARINE POLLUTION BULLETIN 2025年 214卷 117710页

作者： Zhao, Fan Huang, Baoxi Wang, Jiaqi Shao, Xinlei Wu, Qingyang Xi, Dianhan Liu, Yongying Chen, Yijia Zhang, Guochen Ren, Zhiyan Chen, Jundong Mizuno, Katsunori Univ Tokyo Grad Sch Frontier Sci Kashiwa Chiba 2778563 Japan Univ Calif Los Angeles Dept Environm Hlth Sci Los Angeles CA 90095 USA Shiga Univ Data Sci & AI Innovat Res Promot Ctr Hikone Shiga 5228522 Japan

Traditional detection and monitoring of seafloor debris present considerable challenges due to the high costs associated with underwater imaging devices and the complex environmental conditions in marine ecosystems. In response to these challenges, this field study conducted in Koh Tao, Thailand, proposed an innovative and costeffective approach that leverages super-resolution reconstruction (SRR) technology in conjunction with an optimized object detection model based on YOLOv8. Super-resolution (SR) images reconstructed by seven SRR models were fed into the proposed Seafloor-Debris-YOLO (SFD-YOLO) model for seafloor debris object detection. RDN model achieved the highest reconstruction results with a signal-to-noise ratio (PSNR) of 41.02 dB and structural similarity (SSIM) of 95.08 % and attained state-of-the-art (SOTA) accuracy in debris detection with a mean Average Precision (mAP) of 91.2 % using RDN-reconstructed images with a magnification factor of 4. Additionally, this study provided an in-depth analysis of the influence of magnification factors within the SRR process, offering a quantitative comparison of images before and after reconstruction, as well as a comparative evaluation across various detection algorithms with a novel pretraining strategy. This approach to underwater survey methods, combined with SRR technology, marks an advancement in the field of seafloor debris monitoring, presenting practical solutions to enhance image quality affected by field conditions and enabling the precise identification of marine debris.

关键词： Deep learning object detection and classification Pre-training strategy Seafloor debris Super-resolution reconstruction YOLOv8

来源：评论

学校读者我要写书评

暂无评论

Automatic detection of rice disease in images of various leaf sizes

引用

IET IMAGE PROCESSING 2025年第1期19卷

作者： Kiratiratanapruk, Kantip Temniranrat, Pitchayagan Sinthupinyo, Wasin Marukatat, Sanparith Patarapuwadol, Sujin Natl Sci & Technol Dev Agcy NSTDA Natl Elect & Comp Technol Ctr NECTEC Image Proc & Understanding Res Team Pathum Thani Thailand Kasetsart Univ Fac Agr Kamphaeng Saen Dept Plant Pathol Nakhon Pathom Thailand

To help farmers manage limited resources, rice disease diagnosis must be accurate, timely, and affordable. This study addresses challenges in rice field images, such as environmental variability and differences in rice leaf sizes. The proposed technique combines convolutional neural network object detection with image tiling, using estimated rice leaf width as a reference for image division. An 18-layer ResNet model was trained using ground truth leaf width values for regression in leaf width estimation. Experiments used a dataset of 4960 images representing 8 rice diseases. The leaf width prediction model achieved a mean absolute percentage error of 11.18% and was used to generate a tiled dataset for training advanced rice disease detection models. The tiling technique was evaluated using YOLOv4, YOLOv8n, YOLOv8l, DINO-5scale Swin-L, and Co-DINO-5scale Swin-L models by comparing detection performance on original and tiled datasets. Mean average precision improved significantly: YOLOv4 increased from 87.56% to91.14%, YOLOv8n from 89.80% to 91.70%, and YOLOv8l from 89.80% to 93.20%. More advanced models, such as DINO5scale Swin-L and Co-DINO-5scale Swin-L, achieved even higher precision, at 93.40% and 94.20%, respectively. In conclusion, the tiling technique improved detection efficiency and addressed object size variability, enhancing rice disease detection accuracy inreal-world scenarios.

关键词： deep learning object detection and classification relative-size image processing rice disease

来源：评论

学校读者我要写书评

暂无评论

Detecting and classifying rooftops with a CNN-based remote-sensing method for urban area cool roof application

引用

ENERGY REPORTS 2024年 11卷 2516-2525页

作者： Park, Jaehyeong Park, Sangun Kang, Juyoung Ajou Univ Dept Business Analyt Business Sch 206 World Cup Ro Suwon South Korea Kyonggi Univ Coll Software Management Dept Ind & Management Engn 154-42 Gwanggyosan Ro Suwon 15442 South Korea Ajou Univ Dept E Business Business Sch 206 Worldcup Ro Suwon 16499 South Korea

Cool roofs reduce the greenhouse effect and increases the energy efficiency of buildings in urban areas, so they have been continuously researched and developed. Prior cool roof studies measured their efficiency when applied to buildings or objects. However, research on their application is insufficient. Therefore, this study highlights a broader approach to the effectiveness of the cool roof as a key difference from prior research. To do so, the study focuses on revealing the potential benefits of cool roof with practical applicability, by using aerial images to estimate and describe the construction costs and energy savings associated with cool roof construction in a large urban area. This study proceeded as follows. First, aerial images of eight metropolitan cities in South Korea were collected to construct a dataset for remote sensing. A pre -trained Convolutional Neural Network (CNN) model was employed to detect rooftops in each of the images. The detected rooftops were then clustered according to their surface color and their areas were calculated because the current color determines how much energy can be saved by applying cool roofs. Subsequently, a scenario -based cost -benefit analysis was conducted to estimate the benefits of cool roof application. The results show that cool roofs can reduce cooling energy use, thereby reducing greenhouse gas emissions, and increase urban sustainability.

关键词： Cool roof Remote sensing Cooling energy savings object detection and classification

来源：评论

学校读者我要写书评

暂无评论

Riverbed litter monitoring using consumer-grade aerial-aquatic speedy scanner (AASS) and deep learning based super-resolution reconstruction and detection network

引用

MARINE POLLUTION BULLETIN 2024年第Pt A期209卷 117030页

作者： Zhao, Fan Liu, Yongying Wang, Jiaqi Chen, Yijia Xi, Dianhan Shao, Xinlei Tabeta, Shigeru Mizuno, Katsunori Univ Tokyo Grad Sch Frontier Sci Dept Environm Syst Tokyo Japan Univ Tokyo Grad Sch Frontier Sci Dept Socio Cultural Environm Studies Tokyo Japan

Underwater litter is widely spread across aquatic environments such as lakes, rivers, and oceans, significantly impacting natural ecosystems. Current automated monitoring technologies for detecting this litter face limitations in survey efficiency, cost, and environmental conditions, highlighting the need for efficient, consumergrade technologies for automatic detection. This research introduces the Aerial-Aquatic Speedy Scanner (AASS) combined with Super-Resolution Reconstruction (SRR) and an enhanced YOLOv8 detection network. The AASS system boosts data acquisition efficiency over traditional methods, capturing high-resolution images that accurately identify and categorize underwater waste. The SRR technique enhances image quality by mitigating common issues like motion blur and low resolution, thereby improving the YOLOv8 model's detection capabilities. Specifically, the RCAN model achieved the highest mean average precision (mAP) of 78.6 % for object detection accuracy on reconstructed underwater litter among the tested SR models. With a magnification factor of 4, the SR test set shows an improved mAP compared to the Bicubic test set. These results demonstrate the effectiveness of the proposed method in detecting underwater litter.

关键词： Aerial-aquatic speedy scanner Deep-learning object detection and classification Riverbed litter monitoring Super-resolution reconstruction YOLOv8

来源：评论

学校读者我要写书评

暂无评论

Neural Network Technologies for detection and classification of objects

引用

OPTOELECTRONICS INSTRUMENTATION AND DATA PROCESSING 2023年第3期59卷 329-345页

作者： Borzov, S. M. Nezhevenko, E. S. Russian Acad Sci Inst Automat & Electrometry Siberian Branch Novosibirsk 630090 Russia

We present a review of the basic ideas used in solving the problems of detecting and classifying objects by their images using neural network technologies. The key publications on the most popular ways to improve classification accuracy are considered. It is shown that in the last decade, neural network methods for detecting objects have achieved significant success by using convolution technologies and applying deep learning with large databases. The main shortcomings, limitations and possible directions for the improvement of existing approaches are analyzed.

关键词： neural network technologies image processing object detection and classification convolutional neural networks deep learning hybrid techniques

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

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

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：