检索结果-内蒙古大学图书馆

IEEE International Conference on Autonomous Robot Systems and Competitions (ICARSC)

作者： Kasaei, S. Hamidreza Oliveira, Miguel Lim, Gi Hyun Lopes, Luis Seabra Tome, Ana Maria Univ Aveiro IEETA P-3810193 Aveiro Portugal

ISBN: (纸本)9781479942541

Three-dimensional object detection and recognition is increasingly in manipulation and navigation applications in autonomous service robots. It involves clustering points of the point cloud from an unstructured scene into objects candidates and estimating features to recognize the objects under different circumstances such as occlusions and clutter. This paper presents an efficient approach capable of learning and recognizing object categories in an interactive and open-ended manner. In this paper, "open-ended" implies that the set of object categories to be learned is not known in advance. The training instances are extracted from actual experiences of a robot, and thus become gradually available, rather than being available at the beginning of the learning process. This paper focuses on two state-of-the-art questions: (1) How to automatically detect, conceptualize and recognize objects in 3d unstructured scenes in an open-ended manner? (2) How to acquire and utilize high-level knowledge obtained from the user (e. g. category label) to improve the system performance? This approach starts with a pre-processing phase to remove unnecessary information and prepare a suitable point cloud. Clustering is then applied to detect object candidates. Subsequently, all object candidates are described based on a 3d shape descriptor called spin-image. Finally, a nearest-neighbor classification rule is used to assign category labels to the detected objects. To examine the performance of the proposed approach, a leave-one-out cross validation algorithm is utilized to compute precision and recall. The experimental results show the fulfilling performance of this approach on different types of objects.

关键词： open-ended learning 3d object recognition spin-image descriptor autonomous robots

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Two-Stage 3d object recognition from Orthogonal Projections 13

A Hybrid Two-Stage 3D Object Recognition from Orthogonal Pro...

引用

13th IEEE International Conference on Signal Processing and Communication Systems (ICSPCS)

作者： Quang Tri Chiem Lech, Margaret Wilkinson, Richardt H. RMIT Univ Sch Engn Melbourne Vic 3000 Australia

ISBN: (纸本)9781728121949

The availability of large data-sets of 3d scans of objects promoted the potential of applying deep learning techniques to learn how to categorize objects based on their 3d representations automatically. However, a straightforward application of deep learning classifiers to 3d data requires a prohibitive amount of processing power, and therefore, it is not as effective as the classification of objects represented in the 2d domain. This paper address this limitation by mapping the 3d data into 2d representation and applying a hybrid deep and shallow configuration of neural networks to classify different objects. The proposed approach simplifies the network architecture and reduces the computational cost required to train the classifier. The 3d data is mapped into 2d representation using orthogonal projections made from multiple viewpoints. This approach has two additional benefits. Firstly, it increases the number of 2d data samples representing a given object. Secondly, the projected data reduces the occlusion, illumination, and noise problems usually associated with the 3d representation. Given the 2d object representation, the high classification accuracy is achieved through a hybrid, 2-stage classification system. In the first stage, a pre-trained Convolutional Neural Network (CNN) classifier (AlexNet) is fine-tuned to generate vectors of soft labels representing class probabilities. These vectors (or the first-stage CNN features) are passed to a shallow neural network trained to generate the final class labels. While the first stage classifier is trained directly on the physical data, the second stage classifier is trained on the probability vectors (or/and features) of the first stage classifier, and therefore, the classification outcomes of the first stage are in a way corrected by the second stage assessment. Validation experiments confirmed that the second stage classification significantly improves classification outcomes given by the first stage classifier.

关键词： 3d object classification 3d object recognition transfer learning orthogonal projections CNN

来源：评论

学校读者我要写书评

暂无评论

R-COVNET: RECURRENT NEURAL CONVOLUTION NETWORK FOR 3d object recognition 25

R-COVNET: RECURRENT NEURAL CONVOLUTION NETWORK FOR 3D OBJECT...

引用

25th IEEE International Conference on Image Processing (ICIP)

作者： Tchuinkou, danielle Bobda, Christophe Univ Arkansas CSCE Dept Fayetteville AR 72701 USA

ISBN: (纸本)9781479970612

Pointcloud is a very precise digital format for recording objects in space. Pointclouds have received increasing attention lately, due to the higher amount of information it provides compared to images. In this paper, we propose a new deep learning architecture called R-CovNet, designed for 3d object recognition. Unlike previous architectures that usually sample or convert pointcloud into three-dimensional grids before processing, R-CovNet does not require any preprocessing. Our main goal is to provide a permutation invariant architecture specially designed for pointclouds data of any size. Experiments with well-known benchmarks show that R-CovNet can achieve an accuracy of 92.7%, thus outperforming all the volumetric methods.

关键词： RNN Point Cloud 3d object recognition

来源：评论

学校读者我要写书评

暂无评论

New 3d reconstruction approach for 3d object recognition in neuro-vision system

New 3D reconstruction approach for 3D object recognition in ...

引用

Conference on Applications of Artificial Neural Networks in Image Processing III

作者： Xiong, YG Zhang, GZ Zhongshan Univ Dept Radio & Elect Guangzhou 510275 Peoples R China

ISBN: (纸本)0819427470

In this paper, a new 3d reconstruction approach for 3d object recognition in neuro-vision system is presented. First, a phase based stereo matching using Hopfield neural network approach is presented. the stereo matching problems are treated in frequency domain by using local phase, Instead of matching feature or texture of images, the stereo matching process is performed by using local phases of left image and right image in stereo image pair. The Hopfield neural network is adopted to implement the stereo matching process. A suitable architecture of neural network is established, so that the computation can be implemented efficiently in parallel. A suitable matching function is created by using the local phase property. The energy function for neural network is constructed with satisfying some necessary constraints. The stereo matching process then is carried to find the minimum energy corresponding to the solution of the problem. Second, a 3d object reconstruction neural network is constructed by using BP neural network. So the 3d configuration and shape can be reconstructed by this neural network. With multiple neural networks the 3d reconstruction processes can be performed in parallel. The examples for both synthetic and real images are shown in the experiment, and goad results are obtained.

关键词： 3d reconstruction stereo matching neural network image processing local phase 3d object recognition

来源：评论

学校读者我要写书评

暂无评论

NurbsNet: A Nurbs approach for 3d object recognition

NurbsNet: A Nurbs approach for 3d object recognition

引用

International Joint Conference on Neural Networks (IJCNN) held as part of the IEEE World Congress on Computational Intelligence (IEEE WCCI)

作者： Escalona, Felix Viejo, diego Fisher, Robert B. Cazorla, Miguel Univ Alicante Inst Comp Res Alicante Spain Univ Edinburgh Sch Informat Edinburgh Midlothian Scotland

ISBN: (纸本)9781728169262

With the emergence of low cost 3d sensors, the focus is moving towards the recognition and scene understanding of tridimensional data. This kind of representation is really challenging in terms of computation, and it needs the development of new strategies and algorithms to be handled and interpreted. In this work, we propose NurbsNet, a novel approach for 3d object classification based on local similarities with free form surfaces modeled as Nurbs. The proposal has been tested in ModelNetlO and ModelNet40 with results that are promising with less training iterations than state-of-the-art methods and very low memory consumption.

关键词： 3d object recognition neural networks Nurbs

来源：评论

学校读者我要写书评

暂无评论

New 3d reconstruction approach for 3d object recognition in intelligent assembly system

New 3D reconstruction approach for 3D object recognition in ...

引用

SPIE International Conference on Applications and Science of Computational Intelligence

作者： Xiong, YG Zhang, GZ Zhongshan Univ Dept Radio & Elect Canton 510275 Peoples R China

ISBN: (纸本)0819428396

In this paper, a new 3d reconstruction approach for 3d object recognition in neuro-vision system is presented. First, a phase based stereo matching using Hopfield neural network approach is presented. the stereo matching problems are treated in frequency domain by using local phase, Instead of matching feature or texture of images, the stereo matching process is performed by using local phases of left image and right image in stereo image pair. By using the windowed Fourier transform, the windowed Fourier phases can be calculated. Through the variable window Gabor filter, the local phases of image can also be obtained. The Hopfield neural network is adopted to implement the stereo matching process. A suitable architecture of neural network is established, so that the computation can be implemented efficiently in parallel. A suitable matching function is created by using the local phase property. The energy function for neural network is constructed with satisfying some necessary constraints. The stereo matching process then is carried to find the minimum energy corresponding to the solution of the problem. Second, a 3d object reconstruction neural network is constructed by using BP neural network. So the 3d configuration and shape can be reconstructed by this neural network. With multiple neural networks the 3d reconstruction processes can be performed in parallel. The examples for both synthetic and real images are shown in the experiment, and good results are obtained.

关键词： 3d reconstruction stereo matching neural network image processing local phase 3d object recognition

来源：评论

学校读者我要写书评

暂无评论

A fast spin images matching method for 3d object recognition

A fast spin images matching method for 3D object recognition

引用

5th International Symposium on Photoelectronic detection and Imaging - Laser Sensing and Imaging and Applications

作者： Wang Ming-ming Hao Ying-ming Zhu Feng Chinese Acad Sci Shenyang Inst Automat Shenyang 110016 Peoples R China

ISBN: (纸本)9780819497741

Spin image has been applied to 3d object recognition system successfully because of its advantages of rotation, translation and view invariant. However, this method is very time consuming, owning to its high-dimensional characteristics and its complicated matching procedure. To reduce the recognition time, in this paper we propose a coarse-to-fine matching strategy for spin images. There are two steps to follow. Firstly, a low dimensional feature is introduced for a given point. The feature contain two components, its first component is the perpendicular distance from the centroid of the given point's neighbor region to the tangential plane of the given point, its second component is the maximum distance between the projection point of the centroid on the tangential plane and projection points of the neighbor region on the tangential plane. Secondly when comparing a point from a target with a point from a model, their low features are matched first, only if they satisfy the low feature constrains, can they be selected as a candidate point pair and their spin images are further matched by similarity measurement. When all the target points and all the model points finish above matching process, those candidate point pairs with high spin image similarity are selected as corresponding point pairs, and the target can be recognized as the model with the most amount of corresponding point pairs. Experiment based on Stanford 3d models is conducted, and the comparison of experiment results of our method with the standard spin image shows that the propose method is more efficient while still maintain the standard spin image's advantages.

关键词： 3d object recognition Matching Spin image

来源：评论

学校读者我要写书评

暂无评论

View-based 3d object recognition using wavelet multiscale singular-value decomposition and support vector machine

View-based 3D object recognition using wavelet multiscale si...

引用

5th International Conference on Wavelet Analysis and Pattern recognition

作者： Zhai, Jun-Hai Wang, Xi-Zhao Zhang, Su-Fang Li, Jie Hebei Univ Machine Learning Ctr Fac Math & Comp Sci Baoding 071002 Peoples R China Hebei Informat Engn Sch Teaching & Res Sect Math Baoding 071000 Peoples R China

ISBN: (纸本)9781424410651

In this paper, a novel view-based 3d object recognition method is proposed, which consists of three steps. First, employing wavelet transform to decompose view images of the object into different frequency sub-images. Second, for each sub-image, the features are extracted using singular-value decomposition (SVd) approach, and the features extracted from sub-images are combined to construct the feature vector of the original image. Third, the feature vector is fed into the support vector machine (SVM) to classify the objects. Experimental results show that the proposed method is effective.

关键词： 3d object recognition wavelet transform sub-image singular-value decomposition support vector machine

来源：评论

学校读者我要写书评

暂无评论

A novel model for eliminating overlapping issues in 3d object recognition using dove swarm optimization based light GBM

引用

International Journal of Information Technology (Singapore) 2023年第5期15卷 2387-2393页

作者： Rajendran, Ramana Murugan, B.S. Computer Science and Engineering Kalasalingam Academy of Research and Education Tamil Nadu Srivilliputhur India

This paper proposes a novel modified light gradient boosting machine-based dove swarm optimization (MLGBM-dSO) model for the accurate recognition of 3d objects. To evaluate the performance, the proposed MLGBM-dSO model utilizes 3d shape CAd images from the ModelNet dataset as input. The 3d shape images are initially preprocessed through data screening and normalization tasks to produce structured data for the proposed classifier. The preprocessed data is provided to the proposed MLGBM-dSO model that effectively extracts and classifies the 3d object images into distinct categories. Here, the modified gradient boosting machine (GBM) is formed by the integration of a mask region-based convolutional neural network (mask RCNN) and light GBM model. To achieve enhanced object recognition accuracy and the overlapping issues that occur during classification are mitigated by using the dove swarm optimization (dSO) algorithm. The efficiency of the proposed MLGBM-dSO model is evaluated using different evaluation indicators such as accuracy, specificity, recall, intersection over union (IoU), precision, and F1-score. The experimental result inherits that the proposed MLGBM-dSO model achieves higher 3d object recognition accuracy of about 97.2%. © 2023, The Author(s), under exclusive licence to Bharati Vidyapeeth's Institute of Computer Applications and Management.

关键词： 3d object recognition Accuracy dove swarm optimization algorithm Light gradient boosting machine Mask region-based convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

CURVATURE AUGMENTEd dEEP LEARNING FOR 3d object recognition 25

CURVATURE AUGMENTED DEEP LEARNING FOR 3D OBJECT RECOGNITION

引用

25th IEEE International Conference on Image Processing (ICIP)

作者： Braeger, Sarah Foroosh, Hassan UCF Computat Imaging Lab Orlando FL 32816 USA

ISBN: (纸本)9781479970612

This paper presents a new method to incorporate shape information into convolutional neural network (CNN)s for 3d object recognition. Voxel CNNs have been very successful with the task of 3d object recognition. However, continuous shape information that is useful for recognition is often lost in their conversion to a voxel representation. We propose a single dimensional feature that can be applied to voxel CNNs. This paper presents a novel rotation-invariant feature based on mean curvature that improves shape recognition for voxel CNNs. We augment the recent voxel CNN Octnet architecture with our feature and demonstrate a 1% overall accuracy increase on the ModelNet10 dataset.

关键词： 3d object recognition Computational Geometry Convolutional Neural Networks deep Learning

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：