检索结果-内蒙古大学图书馆

作者： dong, Wenbo University of Minnesota

学位级别：Ph.D., Doctor of Philosophy

vision sensors mounted on mobile robotic platforms hold great promise in automated agriculture management. However, established computer vision techniques often fail to perform well in agricultural environments due to the environmental complexity, which makes automation difficult. To address this problem, we have designed and developed three-dimensional (3d) computer vision algorithms that improve the accuracy of imaging devices, suppress the undesirable environmental interferences, and generate accurate and precise 3d models of plants with detailed information automatically extracted for farmers. This dissertation is roughly separated into three main parts. In the first part of the thesis, we study the problem of extrinsic calibration of a 2d laser rangefinder and a camera. We present a novel method for extrinsically calibrating a camera and a 2d laser rangefinder whose beams are invisible from the camera image. We show that the point-to-plane constraints from a single observation of a V-shaped calibration pattern composed of two non-coplanar triangles suffice to uniquely constrain the relative pose between two sensors. We propose an approach to obtain analytical solutions using point-to-plane constraints from single or multiple observations. Along the way, we also show that the previous solutions, in contrast to our method, have inherent ambiguities and therefore must rely on a good initial estimate from a large number of observations. In the second part of the thesis, we study the problem of building coherent 3d reconstructions of orchard rows to improve the accuracy of measuring semantic traits for phenotyping and to automate such measurements. Even though 3d reconstructions of side views can be obtained using standard mapping techniques, merging the two side-views is difficult due to the lack of overlap between the two partial reconstructions. We propose a novel method that utilizes global features and semantic information to obtain an initial solution aligning

关键词： 3d computer vision Algorithm Semantic reconstruction Agricultural environments

来源：评论

学校读者我要写书评

暂无评论

Within-row spacing sensing of maize plants using 3d computer vision

引用

BIOSYSTEMS ENGINEERING 2014年 125卷 54-64页

作者： Nakarmi, Akash d. Tang, Lie Iowa State Univ Dept Agr & Biosyst Engn Ames IA 50010 USA

Within-row plant spacing plays an important role in uniform distribution of water and nutrients among plants which affects the final crop yield. While manual in-field measurements of within-row plant spacing is time and labour intensive, little work has been done on an alternative automated process. We have attempted to develop an automatic system making use of a state-of-the-art 3d vision sensor that accurately measures within-row maize plant spacing. Misidentification of plants caused by low hanging canopies and doubles were reduced by processing multiple consecutive images at a time and selecting the best inter-plant distance calculated. Based on several small scale experiments in real fields, our system has been proven to measure the within-row maize plant spacing with a mean and standard deviation error of 1.60 cm and 2.19 cm, and a root mean squared error of 2.54 cm, respectively. (C) 2014 IAgrE. Published by Elsevier Ltd. All rights reserved.

关键词： Spacing sensing Inter-plant spacing Within-row 3d computer vision Time-of-flight

来源：评论

学校读者我要写书评

暂无评论

Robot Learning from demonstration Using 3d computer vision

Robot Learning from Demonstration Using 3D Computer Vision

引用

International Conference on Advanced Material and Manufacturing Science (ICAMMS 2012)

作者： debono, James Aaron Fang, Gu Univ Western Sydney Sch Comp Engn & Math Sydney NSW 2751 Australia

ISBN: (纸本)9783037859933

For robot application to proliferate in industry, and in unregulated environments, a simple means of programming is required. This paper describes methods for robot Learning from demonstration (Lfd). These methods used an RGB-d sensor for demonstration observation, and used finite state machines (FSMs) for policy derivation. Particularly, a method for object recognition was developed, which required only a single frame of data for training, and was able to perform real-time recognition. A planning method for object grasping was also developed. Experiments with a pick-and-place robot show that the developed methods resulted in object recognition accuracy greater than 99% in cluttered scenes, and manipulation accuracies of below 3mm in linear motion and 2 in rotation.

关键词： learning from demonstration 3d computer vision automated manufacturing

来源：评论

学校读者我要写书评

暂无评论

3d computer vision on the Android Platform

3D Computer Vision on the Android Platform

引用

作者： Vášová, Viktorie Charles University of Prague

The aim of this work is to study the problem of automatically reconstructing a 3d scene from a pair of its photographs. It investigates to what extent can this task be performed on a mobile phone running on the Android platform. Several image correspondence algorithms are discussed. The specifics of the Android platform from developers perspective are also considered. The algorithms are then utilized in a particular way to obtain a mobile software solution for 3d reconstruction. This software runs solely on a mobile phone, including capturing the photos by the phones camera and the final visualization. An implementation of the application is a part of this work.

关键词： 3d computer vision image correspondence Android platform

来源：评论

学校读者我要写书评

暂无评论

Optimal triangulation in 3d computer vision using a multi-objective evolutionary algorithm

引用

EvoWorkshops 2007

作者： Vite-Silva, Israel Cruz-Cortes, Nareli Toscano-Pulido, Gregorio de la Fraga, Luis G. CINVESTAV Dept Comp Av IPN 2508 Mexico City 73060 DF Mexico CINVESTAV Unidad Tamaulipas Tamps 87276 Mexico

ISBN: (纸本)9783540718048

The triangulation is a process by which the 3d point position can be calculated from two images where that point is visible. This process requires the intersection of two known lines in the space. However, in the presence of noise this intersection does not occur, then it is necessary to estimate the best approximation. One option towards achieving this goal is the usage of evolutionary algorithms. In general, evolutionary algorithms are very robust optimization techniques, however in some cases, they could have some troubles finding the global optimum getting trapped in a local optimum. To overcome this situation some authors suggested removing the local optima in the search space by means of a single-objective problem to a multi-objective transformation. This process is called multi-objectivization. In this paper we successfully apply this multi-objectivization to the triangulation problem.

关键词： evolutionary multi-objective optimization 3d computer vision triangulation

来源：评论

学校读者我要写书评

暂无评论

Ground truth evaluation of 3d computer vision on non-rigid biological structures

Ground truth evaluation of 3D computer vision on non-rigid b...

引用

5th European Conference on Precision Agriculture/2nd European Conference Precision Livestock Farming

作者： Nielsen, M Andersen, HJ Slaughter, dC Granum, E Aalborg Univ Aalborg Denmark

ISBN: (纸本)9076998698

There is an increasing interest in using 3d computer vision in precision agriculture. This calls for better quantitative evaluation and understanding of computer vision methods. It is difficult to evaluate the performance of an algorithm quantitatively, when dense ground truth such as dense depth map is not readily available, in contrast to sparse ground truth data at high levels of abstraction. This paper proposes a test framework using ray traced crop scenes. This enables quantitative comparison of dense 3d reconstruction algorithms, camera- and lighting setups, and post-processing before investing in expensive equipment and field experiments. In order to illustrate the possibilities in this test framework, 3240 disparity maps were computed and example hypotheses were analysed using statistics and graphs. It was to be found a valuable tool to structure the otherwise incomprehensible vast information space and to see relationships between parameter configurations and crop features.

关键词： 3d computer vision benclunarking ray-tracing remote sensing

来源：评论

学校读者我要写书评

暂无评论

Robot Learning from demonstration Using 3d computer vision

Robot Learning from Demonstration Using 3D Computer Vision

引用

The 2012 International Conference on Advanced Material and Manufacturing Science(ICAMMS 2012)

作者： James Aaron debono Gu Fang School of Computing Engineering&MathematicsUniversity of Western Sydney

For robot application to proliferate in industry,and in unregulated environments,a simple means of programming is requiredThis paper describes methods for robot Learning from demonstration(LAd)These methods used an RGB-d sensor for demonstration observation,and used finite state machines(FSMs) for policy derivationParticularly,a method for object recognition was developed,which required only a single frame of data for training,and was able to perform real-time recognitionA planning method for object grasping was also developedExperiments with a pick-and-place robot show that the developed methods resulted in object recognition accuracy greater than 99%in cluttered scenes,and manipulation accuracies of below 3mm in linear motion and 2° in rotation

关键词： learning from demonstration 3d computer vision automated manufacturing

来源：评论

学校读者我要写书评

暂无评论

design and highly accurate 3d displacement characterisation of monolithic SMA micro-gripper using computer vision

Design and highly accurate 3D displacement characterisation ...

引用

Conference on Microrobotics and Microsystem Fabrication

作者： Bellouard, Y Sulzmann, A Jacot, J Clavel, R Ecole Polytech Fed Lausanne Dept Microtech Inst Microtech CH-1015 Lausanne Switzerland

ISBN: (纸本)0819426342

In the robotics field, several grippers have been developed using SMA technologies, but, so far, SMA is only used as the actuating part of the mechanical device. However mechanical device requires assembly and in some cases this means friction. In the case of micro-grippers, this becomes a major problem due to the small size of the components. In this paper, a new monolithic concept of micro-gripper is presented. This concept is applied to the grasping of sub-millimetre optical elements such as Selfoc(TM) lenses and the fastening of optical fibres. Measurements are performed using a newly developed high precision 3d-computer vision tracking system to characterise the spatial positions of the micro-gripper in action. To characterise relative motion of the micro-gripper the natural texture of the micro-gripper is used to compute 3d displacement. The microscope image CCd receives high frequency changes in light intensity from the surface of the gripper. Using high resolution camera calibration, passive auto focus algorithms and 2d object recognition, the position of the micro-gripper can be characterised in the 3d workspace and can be guided in future micro assembly tasks.

关键词： micro gripper shape memory alloys 3d microscopy 3d computer vision micro actuator characterisation

来源：评论

学校读者我要写书评

暂无评论

Introducing 3d vision and computer graphics to archaeological workflow -: An applicable framework

Introducing 3D vision and computer graphics to archaeologica...

引用

3rd International Conference on computer vision Theory and Applications

作者： Mara, Hubert Monitzer, Andreas Stoettinger, Julian Univ Florence VAST Lab PIN Serv Didatt & Sci Piazza Ciardi 25 I-59100 Prato Italy Vienna Univ Technol Inst Comp Graph & Algorithms A-1040 Vienna Austria Vienna Univ Technol Inst Comp Aided Automat Pattern Recognit & Image Proc Grp A-1040 Vienna Austria

ISBN: (纸本)9789898111210

Cataloging drawings of ancient vessels and sherds is still the most time consuming task in the typical archaeological workflow. The properties of these findings like profile, volume, and wall thickness have always been estimated and drawn by hand. Through archiving, classifying and exhibiting these ancient artifacts we wish to gather as precise information as possible. Within seconds, today's 3d-scanners provide surface meshes of ancient vessels which are more precise than any manual estimation which may take up to several hours. We propose a semi-automated, applicable framework for dealing with large 3d-meshes of ancient findings from scanning the vessels for publication. In this interactive environment we estimate the axis of vessels, estimate their profile lines and render real time visualizations using state-of-the-art 3d-hardware techniques. The results can be printed in their real size for direct use in archaeological literature. Further, these methods will give the ability to publish 3d-meshes of ancient vessels for archaeological research. Recent extended tests have been carried out on archaeological sites in Peru and Austria. These experiments showed under real life circumstances the improvement of using this system in both precision and time efficiency.

关键词： 3d computer vision bi-directional reflectance distribution function (BRdF) rotational axis cultural heritage archaeology

来源：评论

学校读者我要写书评

暂无评论

Hybrid Method for Point Cloud Classification

引用

IEEE ACCESS 2025年 13卷 8825-8838页

作者： Hazer, Abdurrahman Yildirim, Remzi Ankara Yildirim Beyazit Univ Grad Sch Nat & Appl Sci TR-06010 Ankara Turkiye Tokat Gaziosmanpasa Univ Dept Comp Engn TR-60250 Ankara Turkiye

In this study, we introduce a novel hybrid ResPOL method that integrates a bidirectional architecture with a newly developed Patch-Offset Lambda (POL) mechanism for feature extraction from point cloud data. This hybrid approach effectively combines the Residual Machine Learning Perceptron (ResMLP) with the POL mechanism to capture both local and global features of 3d point clouds. The ResMLP component hierarchically extracts high-frequency local features from point cloud patches, while the POL mechanism is specifically designed to capture low-frequency global features. This dual extraction process ensures minimal loss of both high and low-frequency features. The POL mechanism employs Lambda layers within a linear framework, significantly enhancing the classification speed and accuracy compared to traditional attention mechanisms that suffer from quadratic complexity and non-linear structures. By processing local and global features in parallel, the Hybrid ResPOL method combines these features and feeds them into the classification head, optimizing performance. Experimental results indicate that the Hybrid ResPOL method achieves an overall accuracy (OA) of 94.3% and a mean accuracy (mAcc) of 91.3% on the ModelNet40 dataset. Additionally, it demonstrates robust performance on the challenging ScanObjectNN dataset, with accuracies of 92% OA and 91.2% mAcc. The method processes data at rates of 205.1 samples per second during training and 493.6 samples per second during testing, outperforming the PointMLP method by a factor of 4.3. The superior performance of Hybrid ResPOL in both accuracy and speed highlights its effectiveness over existing attention-based methods.

关键词： Point cloud classification patch offset lambda linear attention deep learning 3d computer vision Point cloud classification patch offset lambda linear attention deep learning 3d computer vision

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：