检索结果-内蒙古大学图书馆

A deep learning system for collotelinite segmentation and coal reflectance determination

INTERNATIONAL JOURNAL OF COAL GEOLOGY 2022年 263卷

作者： Magalhaes Santos, Richard Bryan Augusto, Karen Soares Alvarez Iglesias, Julio Cesar Rodrigues, Sandra Paciornik, Sidnei Esterle, Joan S. Alcantara Domingues, Alei Leite Pontifical Catholic Univ Rio de Janeiro DEQM PUC Dept Chem & Mat Engn Rua Marques de Sao Vicente 225 BR-22451900 Rio De Janeiro Brazil Deloitte Touche Tohmatsu Artificial Intelligence & Data Rio De Janeiro Brazil Univ Queensland Sch Earth & Environm Sci St Lucia Qld 4072 Australia CTF VALE Ave Dr Marco Paulo Simon Jardim 3580 BR-34006270 Nova Lima MG Brazil

Coal is widely used in industrial applications such as carbonization and coke production for steel making, combustion, or gasification to generate electricity, and liquefaction to generate petrochemical feedstock. The utility of a coal is dictated by its properties, commonly referred to as its rank, type, and grade. Coal type or maceral composition and coal rank determination by vitrinite (collotelinite) reflectance are traditionally con-ducted manually by trained petrographers using reflected light optical microscopy and bulk chemical tests. This study developed an automatic method based on machine learning for rank determination on petrographic images of coal that might improve the efficiency of this process and potentially eliminate operator subjectivity. Firstly, a Mask R-CNN-based architecture deep learning approach was developed to identify and segment the collotelinite maceral, which is fundamental for rank analysis, as rank can be assessed by the reflectance of this maceral. Secondly, an image processing method was developed to analyze the collotelinite segmented images and determine coal rank by associating the grey values with the reflectance. For the segmentation, five samples were used to train the network, 174 images were used for training, and 86 were used in the test set, with over 80% success rates. Four of those five samples were used alongside another eight to determine the rank. The samples ranged from 0.97% to 1.8% reflectance. The root mean square error calculated between the proposed method and the reference values of reflectance was 0.0978%.

关键词： Coal petrography Vitrinite maceral group Coal rank Computer vision Deep learning Instance segmentation Optical microscopy

来源：评论

学校读者我要写书评

暂无评论

Research on Korla Fragrant Pear Grading Method Based on Random Forest and MLP

Research on Korla Fragrant Pear Grading Method Based on Rand...

引用

image processing and Computer applications (ICIPCA), IEEE International Conference on

作者： Yingchao Wang Jiangyu Zhang Na Li De'an Huang Lihao Qin Shan Sun Xingxing Xie Xinjiang University of Science and Technology Xinjiang Korla China

ISBN: (数字)9798350360240

ISBN: (纸本)9798350384161

In response to the problems of uneven quality and unstable accuracy caused by manual grading in the grading process of Korla pear, a Korla pear grading method based on random forest and MLP is proposed, aiming to further optimize the above problems using machine vision methods. This paper collects left and right views and top views of pear images. Using a combination of K-means clustering and threshold segmentation to achieve background separation; This paper constructs the minimum bounding rectangle for the left and top views of Xiangli to obtain its longitudinal and transverse diameters; Based on the measurement difficulty of pear weight, a pear weight prediction model based on random forest model is constructed using pear diameter parameters; This paper constructs a pear grading model using MLP neural network. The experimental results show that the proposed Korla pear grading method based on random forest and MLP has an accuracy of 99.69% in the training set and 98.75 % in the test set, verifying the feasibility and accuracy of this method.

关键词： Weight measurement Training image segmentation Accuracy Computational modeling Neural networks Object segmentation Manuals Predictive models Random forests

来源：评论

学校读者我要写书评

暂无评论

3D Reconstruction by Pretrained Features and Visual-Inertial Odometry

3D Reconstruction by Pretrained Features and Visual-Inertial...

引用

Asia-Pacific International Symposium on Aerospace Technology (APISAT)

作者： Kunbum, Park Tsuchiya, Takeshi Univ Tokyo Grad Sch Engn Aeronaut & Astronaut Tokyo Japan Univ Tokyo Bunkyou Ku Hongo Tokyo 1138656 Japan

ISBN: (纸本)9789811926358;9789811926341

The goal of my paper is to create a new framework which provides sufficient semantic information for decision makings as a core component of applications such as SLAM (Simultaneous Localization And Mapping), robotics, AR (Augmented Reality), autonomous driving, etc. This framework does not provide dense point clouds. Rather, a scene is described with several features which the agent has been trained to recognize. Specifically, scenes are generated by extracting features from the space's occupancy, the location of the light source, shape of the object, colors, and textures in the images. The extraction of the features is conducted by ensemble with deep-learning based feature extractors, traditional machine learning algorithms and image processing techniques. The deep-learning based feature extractors are trained by a supervised learning with data augmentation over 3D models;and the results of inferences are evaluated by a depth camera and retrained by unsupervised learning. Using existed methods it is difficult to utilize semantic information for unknown objects. But the agent in this methodology tries to describe them as much as possible by utilizing information trained in advance. For an odometry module, which estimates attitudes and positions, is implemented by a typical feature-based visual odometry methodology. The camera coordinate frame's depth camera points and the pixel plane's points are optimized by Levenberg-Marquardt algorithm after extracting a typical corner detection algorithm and tracking by an optical flow algorithm. Using several key-frames, sliding-windowed PnP (Perspective-n-Point), algorithms can be constructed. The visual odometry of the camera and the attitude estimation from the IMU (Inertial Measurement Unit) are loosely coupled. An ARS (Attitude Reference System) is built with a quaternion based linear Kalman filter, and mainly compensates the rotation error of the sliding-windowed PnP algorithm for each frame. The positions of the recogniz

关键词： SLAM Visual-inertial odometry Scene reconstruction Scene understanding Ensemble vision machine learning

来源：评论

学校读者我要写书评

暂无评论

Multi-Metric Induced Robust Graph

引用

IEEE ACCESS 2021年 9卷 94996-95005页

作者： Geng, Zengmin Sun, Xiaodong Du, Jianxia Zhang, Jujian Zhou, Yiling Beijing Inst Fash Technol Div Basic Courses Beijing 100029 Peoples R China Beijing Inst Fash Technol Sch Fash Beijing 100029 Peoples R China

Graph-based learning model has a wide range of applications in machine learning and computer vision. The key issue of the graph-based applications is to construct an informative graph to effectively represent data correlations. In practice, real-world data is usually contaminated by complex noise beyond Gaussian noise and sparse noise, which degrades learning performance dramatically. To construct a robust graph that represents real-world data distribution well, we propose a novel graph construction method. The proposed method is designed to be robust to complex noise beyond sparse or Gaussian noise by combining different previously given metrics of measuring data errors, which robustly and correctly generates a set of edges. To demonstrate the superiority of the proposed method, we employed the clustering and classification tasks on the real-world datasets. The experimental results demonstrate that the proposed method is superior to the existing methods in both accuracy and robustness to complex noise.

关键词： Measurement Noise measurement Measurement uncertainty Gaussian noise Error analysis Correlation Estimation Graph robust image processing noise

来源：评论

学校读者我要写书评

暂无评论

Human-Centric Multimodal machine Learning: Recent Advances and Testbed on AI-Based Recruitment

引用

SN Computer Science 2023年第5期4卷 434页

作者： Peña, Alejandro Serna, Ignacio Morales, Aythami Fierrez, Julian Ortega, Alfonso Herrarte, Ainhoa Alcantara, Manuel Ortega-Garcia, Javier Universidad Autónoma de Madrid Madrid 28049 Spain

The presence of decision-making algorithms in society is rapidly increasing nowadays, while concerns about their transparency and the possibility of these algorithms becoming new sources of discrimination are arising. There is a certain consensus about the need to develop AI applications with a Human-Centric approach. Human-Centric machine Learning needs to be developed based on four main requirements: (i) utility and social good;(ii) privacy and data ownership;(iii) transparency and accountability;and (iv) fairness in AI-driven decision-making processes. All these four Human-Centric requirements are closely related to each other. With the aim of studying how current multimodal algorithms based on heterogeneous sources of information are affected by sensitive elements and inner biases in the data, we propose a fictitious case study focused on automated recruitment: FairCVtest. We train automatic recruitment algorithms using a set of multimodal synthetic profiles including image, text, and structured data, which are consciously scored with gender and racial biases. FairCVtest shows the capacity of the Artificial Intelligence (AI) behind automatic recruitment tools built this way (a common practice in many other application scenarios beyond recruitment) to extract sensitive information from unstructured data and exploit it in combination to data biases in undesirable (unfair) ways. We present an overview of recent works developing techniques capable of removing sensitive information and biases from the decision-making process of deep learning architectures, as well as commonly used databases for fairness research in AI. We demonstrate how learning approaches developed to guarantee privacy in latent spaces can lead to unbiased and fair automatic decision-making process. Our methodology and results show how to generate fairer AI-based tools in general, and in particular fairer automated recruitment systems. © 2023, The Author(s).

关键词： Automated recruitment Bias Biometrics Computer vision Deep learning FairCV Fairness Multimodal Natural language processing

来源：评论

学校读者我要写书评

暂无评论

Handwriting Recognition System Using YOLO and CTC

Handwriting Recognition System Using YOLO and CTC

引用

2023 International Conference on Advanced Computing and Communication Technologies, ICACCTech 2023

作者： Garg, Neeraj Sharma, Neelam Jain, Gautam Jain, Vipul Upreti, Vaibhav Maharaja Agrasen Institute of Technology Department of Artificial Intelligence and Machine Learning India

ISBN: (纸本)9798350380880

In a multitude of applications, handwriting recognition holds significant importance, including tasks like transforming handwritten documents into digital format, automating form processing, and intelligent character recognition. This paper introduces a novel handwriting recognition system that integrates the You Only Look Once (YOLO) object detection algorithm with the Simple Handwritten Text Recognition (HTR) model. The YOLO algorithm is widely recognized for its ability to perform real-time object detection. By adapting it to the task of handwriting recognition, we can detect and localize handwritten text regions within images. The YOLO algorithm is trained on a large dataset of annotated images, enabling it to learn features specific to handwriting. Once the text regions are detected, the HTR model is employed for accurate recognition of the detected text. HTR is a deep learning-based approach that leverages recurrent neural networks (RNNs) to transcribe handwritten text into machine-readable format. The model is trained on a diverse set of labeled handwriting samples, enabling it to effectively recognize and convert handwritten characters into digital text. Our proposed system combines the strengths of YOLO's robust object detection capabilities with HTR's accurate text recognition. The integration of these two components allows for efficient and reliable handwriting recognition, even in challenging scenarios with varying handwriting styles and image qualities. To evaluate the performance of our system, we conducted experiments on benchmark datasets and compared the results with existing approaches. The experimental results demonstrate that our system achieves superior performance in terms of both detection accuracy and recognition accuracy, surpassing state-of-the-art methods. Overall, our proposed handwriting recognition system utilizing YOLO and HTR presents a robust and effective solution for converting handwritten text into digital format. This system can

关键词： Computer vision

来源：评论

学校读者我要写书评

暂无评论

Evaluation of the quantization error in convergence stereo cameras

引用

JOURNAL OF OPTICAL TECHNOLOGY 2020年第8期87卷 495-500页

作者： Sayyedbarzani, S. A. Emam, S. M. Birjand Univ Technol Dept Mech Engn Birjand Iran

Computer vision techniques including image acquisition, image processing, and data analysis are currently popular and widespread. Stereo vision systems in the area of computer vision have been widely used in many applications, such as 3D depth reconstruction, 3D gaming, robot vision navigation, remote sensing, medical image processing, and so on. However, highly accurate 3D stereo vision scanners do not completely satisfy industry requirements due to the high price of these systems. In this paper, what we believe to be a novel method is proposed that increases the 3D measurement accuracy of stereo vision systems with low resolution and cheap cameras. This was achieved through discussing the quantization error of the 3D depth reconstruction of stereo cameras. Accordingly, a mathematical model was introduced to evaluate the quantization error of convergence stereo cameras. Meanwhile, a synthetic experiment was employed to validate our proposed mathematical model. Finally, simulation experiments were conducted to demonstrate how to apply the presented approach to increase the accuracy measurement of low resolution stereo vision cameras. The simulation results show a considerable measurement improvement by changing the convergence angle of stereo vision cameras. (C) 2020 Optical Society of America

关键词： image processing image reconstruction machine vision Medical image processing Remote sensing Three dimensional reconstruction

来源：评论

学校读者我要写书评

暂无评论

Review on the Application of machine vision Algorithms in Fruit Grading Systems 5th

Review on the Application of Machine Vision Algorithms in Fr...

引用

5th International Conference on Intelligent, Interactive Systems and applications, IISA 2020

作者： Kaiyan, Lin Chang, Liu Huiping, Si Junhui, Wu Jie, Chen Modern Agricultural Science and Engineering Institute Tongji University Shanghai China

ISBN: (纸本)9783030637835

China is the largest fruit producer in the world, so fruit grading is one of important tasks in agricultural production. Because manual testing is time-consuming and laborious, it is reasonable to be replaced by machine vision technology. This review included three parts. The first part introduced the architecture of the fruit grading system, including image grasp and processing algorithms. And the system transferred images to computer for processing after captured by acquisition sensor. Then computer performed image processing operations, including preprocessing, segmentation, feature extraction and selection. The second part summarized the machine vision classification algorithms for fruit grading. Classification algorithms are classified as supervised machine learning, unsupervised algorithms, and deep neural network. The supervised machine learning algorithms are Naive Bayes, K-nearest Neighbor, Support Vector machine and so on. The unsupervised machine learning algorithms included K-means Clustering and Principal Component Analysis. And the deep learning part mainly introduced Artificial Neural Network and Convolutional Neural Network. This part reviewed the characteristics and application potential of the above classification algorithms, and comprehensively compared the merits and demerits of assorted classification algorithms applied in fruit classification. The third one discussed the difficulties and challenges of fruit grading tasks based on machine vision technology and put forward some prospects. Finally, this paper summarized and analyzed the machine vision algorithms in fruit grading systems and concluded that the research of machine learning and deep learning combined with machine vision systems is the main development trend in the field of fruit grading and will be a hot research spot in future. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Fruits

来源：评论

学校读者我要写书评

暂无评论

A Real-Time Waste Segregation Using machine vision for Recycling Facilities

A Real-Time Waste Segregation Using Machine Vision for Recyc...

引用

China International Electrical and Energy Conference (CIEEC)

作者： Muhammad Waqas Qaisar Jingyang Fang Umar Masood School of Control Science and Engineering Shandong University Jinan China School of Electrical Engineering Namal University Mianwali Pakistan

ISBN: (数字)9798350359558

ISBN: (纸本)9798350359565

Recycling facilities are often challenged by the demanding task of waste segregation, a process traditionally requiring substantial time and manpower, and lacking efficient automated solutions. Addressing this, our study explores the use of convolutional neural networks (CNNs), particularly the ResNet-18 model, known for its balance of accuracy and computational efficiency compared to its deeper counterparts like ResNet-50 or ResNet-101. ResNet-18 provides sufficient complexity for precise image classification while enabling faster processing speeds essential for real-time applications. Our method employs this CNN to effectively categorize waste into five distinct types, streamlining the sorting process. In this paper, we introduce a custom-designed hardware system, incorporating motor control for the physical segregation of waste. This system operates in tandem with the ResNet-18's classification results, ensuring that each type of waste is accurately sorted into designated bins. Our integrated approach, which pairs the image classification capabilities of ResNet-18 with a practical hardware automation system, aims to improve the efficiency and accuracy of waste segregation in recycling facilities.

关键词： Waste management Accuracy machine vision Real-time systems Hardware Recycling Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Data Storage and Management for image AI Pipelines 25

Data Storage and Management for Image AI Pipelines

引用

Companion of the 2025 International Conference on Management of Data

作者： Utku Sirin Stratos Idreos Harvard University Boston MA USA

ISBN: (纸本)9798400715648

image AI is essential for various applications, such as self-driving cars, medical imaging, and smart farming. Data management is key for efficient image AI, from how to store images to how to manage data while processing the images. This tutorial overviews the emerging area of image AI pipelines by combining approaches from various cross-disciplinary areas such as data management, digital signal processing, computer vision, and machine learning. We specifically focus on image storage and data *** tutorial first gives an overview of image AI pipelines step by step, how they work and the main the challenges. We then describe the main approaches to making image AI pipelines more efficient. We first cover how image AI pipelines store images based on standard storage formats, learned formats, task-specific learned formats, and self-designed formats. Second, we cover how state-of-the-art approaches manage data within image AI pipelines. We identify and describe three main approaches to making image AI pipelines more efficient by efficiently managing data within the pipeline: (i) compressing intermediate data, (ii) materializing and re-using data objects, and (iii) parallelism for better hardware utilization. Lastly, the tutorial covers open data management and systems problems and future directions in making image AI pipelines more efficient.

关键词： ai

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：