检索结果-内蒙古大学图书馆

Adaptive scene dependent filters for segmentation and online learning of visual objects

NEUROCOMPUTING 2007年第7-9期70卷 1235-1246页

作者： Steil, J. J. Goetting, M. Wersing, H. Koerner, E. Ritter, H. Univ Bielefeld Fac Technol Neuroinformat Grp D-33501 Bielefeld Germany Honda Res Inst GmbH D-63073 Offenbach Germany

We propose the adaptive scene dependent filter (ASDF) hierarchy for unsupervised learning of image segmentation, which integrates several processing pathways into a flexible, highly dynamic, and real-time capable vision architecture. It is based on forming a combined feature space from basic feature maps like, color, disparity, and pixel position. To guarantee real-time performance, we apply an enhanced vector quantization method to partition this feature space. The learned codebook defines corresponding best-match segments for each prototype and yields an over-segmentation of the object and the surround. The segments are recombined into a final object segmentation mask based on a relevance map, which encodes a coarse bottom-up hypothesis where the object is located in the image. We apply the ASDF hierarchy for preprocessing input images in a feature-based biologically motivated object recognition learning architecture and show experiments with this real-time vision system running at 6 Hz including the online learning of the segmentation. Because interaction with user is not perfect, the real-world system acquires useful views effectively only at about 1.5 Hz, but we show that for training a new object one hundred views taking only one minute of interaction time is sufficient. (c) 2007 Elsevier B.V. All rights reserved.

关键词： visual online learning unsupervised image segmentation vector quantization cognitive vision object recognition human-machine interaction

来源：评论

学校读者我要写书评

暂无评论

Sparse depth densification for monocular depth estimation

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2024年第5期83卷 14821-14838页

作者： Liang, Zhen Fang, Tiyu Hu, Yanzhu Wang, Yingjian Beijing Univ Posts & Telecommun Beijing Key Lab Work Safety Intelligent Monitoring Beijing 100876 Peoples R China Beijing Univ Posts & Telecommun Sch Modern Post Sch Automat Beijing 100876 Peoples R China Shandong Univ Sch Control Sci & Engn Jinan 250100 Peoples R China

Now the dense depth prediction by single image and a few sparse depth measurements has attracted more and more attention because it provides a low-cost and efficient solution for estimating high-quality depth information. But the current existing methods for the field only take sparse depth as an independent dimension, and the relationship between sparse depth and image itself is always ignored, which undoubtedly limits the improvement of prediction accuracy. For solving the problem, in this paper, a sparse depth densification method is proposed to fully mine the relationship between sparse depth and image for achieving more accurate depth estimation. Based on a priori that the object areas of same category have similar depth values, a Depth Densification Map (DDM) is constructed by the segmentation label obtained from unsupervised image segmentation and sparse depth to realize sparse depth densification. Meantime, considering the potential error of DDM, a Depth Error Map (DEM) is designed to further correct DDM. Then, we use the idea of multi-scale fusion to build a depth estimation network. Finally, the proposed maps are combined with single image as the input of the network, and used to carry out the actual training and testing. Extensive experiments on NYU Depth v2 and Make3D datasets demonstrate the superiority of our proposed approach. Our code is available at . .

关键词： Monocular depth estimation unsupervised image segmentation Sparse depth densification Multi-scale fusion

来源：评论

学校读者我要写书评

暂无评论

unsupervised Surface Reflectance Field Multi-segmenter 16th

Unsupervised Surface Reflectance Field Multi-segmenter

引用

16th International Conference on Computer Analysis of images and Patterns (CAIP)

作者： Haindl, Michal Mikes, Stanislav Kudo, Mineichi Czech Acad Sci Inst Informat Theory & Automat Prague Czech Republic Hokkaido Univ Grad Sch Engn Sapporo Hokkaido Japan

ISBN: (纸本)9783319231921;9783319231914

An unsupervised, illumination invariant, multi-spectral, multi-resolution, multiple-segmenter for textured images with unknown number of classes is presented. The segmenter is based on a weighted combination of several unsupervised segmentation results, each in different resolution, using the modified sum rule. Multi-spectral textured image mosaics are locally represented by eight causal directional multi-spectral random field models recursively evaluated for each pixel. The single-resolution segmentation part of the algorithm is based on the underlying Gaussian mixture model and starts with an over segmented initial estimation which is adaptively modified until the optimal number of homogeneous texture segments is reached. The performance of the presented method is extensively tested on the Prague segmentation benchmark both on the surface reflectance field textures as well as on the static colour textures using the commonest segmentation criteria and compares favourably with several leading alternative image segmentation methods.

关键词： unsupervised image segmentation Textural features Illumination invariants Surface reflectance field Bidirectional texture function

来源：评论

学校读者我要写书评

暂无评论

ILLUMINATION INVARIANT unsupervised SEGMENTER

ILLUMINATION INVARIANT UNSUPERVISED SEGMENTER

引用

16th IEEE International Conference on image Processing

作者： Haindl, Michal Mikes, Stanislav Vacha, Pavel ASCR Inst Informat Theory & Automat Prague 18208 Czech Republic

ISBN: (纸本)9781424456536

A novel illumination invariant unsupervised multispectral texture segmentation method with unknown number of classes is presented. Multispectral texture mosaics are locally represented by illumination invariants derived from four directional causal multispectral Markovian models recursively evaluated for each pixel. Resulted parametric space is segmented using a Gaussian mixture model based unsupervised segmenter. The segmentation algorithm starts with an over segmented initial estimation which is adaptively modified until the optimal number of homogeneous texture segments is reached. The performance of the presented method is extensively tested on the large illumination invariant benchmark from the Prague segmentation Benchmark using 21 segmentation criteria and compares favourably with an alternative segmentation method.

关键词： unsupervised image segmentation illumination invariance

来源：评论

学校读者我要写书评

暂无评论

An In-situ Real-time Hidden Damage Inspection on C-17 Globemaster III Composite Aileron using LSP Technique under Thermal Excitation

An In-situ Real-time Hidden Damage Inspection on C-17 Globem...

引用

Conference on Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems

作者： Fong, Rey-Yie Yuan, Fuh-Gwo North Carolina State Univ Dept Mech & Aerosp Engn Smart Struct & Mat Lab Raleigh NC 27695 USA NIA Integrated Struct Hlth Management Lab Hampton VA 23666 USA

ISBN: (数字)9781510640122

ISBN: (纸本)9781510640122

A non-contact, full-field vision-based non-destructive inspection (V-NDI) system was developed with multiple damages detection capabilities in composite structures under thermal excitation. In contrast to point-based non-destructive inspection (P-NDI) systems employing laser Doppler vibrometer (LDV) with discrete wavefield captured by pointwise scanning, the V-NDI system captures higher spatial resolution wavefield by a CMOS camera for every time instance without repeating the experiment tens of thousand times to reassemble the wavefield like P-NDI. An Advanced Damage Processing Network (ADPNet) was proposed with laser speckle photometry (LSP) employed in a V-NDI system for hidden damages inspection. The LSP/ADPNet system relies on observing the variation of speckle clouds in time sequence without a baseline and is very insensitive to ambient noise with statistics based image processing where traditional holography/ESPI suffers greatly. Other advances of LSP/ADPNet system are its robust tolerance of laser coherence, larger illumination area, flexible choice of correlation functions, and more advanced post-processing techniques such as Bayesian updating/inference or unsupervised image segmentation that can be readily applied. Thermal excitation can have very large power throughput in hundreds of watts compared to traditional PZT actuator, merely in a few watt ranges. Laser speckle itself is the result of self-interference scatter field reflected from a rough surface, and each speckle can be treated as a sensing point from a randomly distributed speckle cloud. By observing the variation of speckle cloud on the structure surface, displacement related quantities in higher dimensions (e.g. hypercomplex envelope, phase between real-valued signal and its quadrature, phase congruency, etc.) can be deduced by the Riesz bp transform and correlated in time sequence to highlight the location of hidden damages in a very effective way. Another novelty of this paper is to make a s

关键词： BVID non-destructive inspection LSP ADPNet Riesz bp transform hypercomplex domain unsupervised image segmentation FPGA wavefield reconstruction DIC

来源：评论

学校读者我要写书评

暂无评论

unsupervised Camouflaged Object segmentation as Domain Adaptation

Unsupervised Camouflaged Object Segmentation as Domain Adapt...

引用

IEEE/CVF International Conference on Computer Vision (ICCV)

作者： Zhang, Yi Wu, Chengyi Ecole Technol Super LIVIA Montreal PQ Canada Henan Polytech Univ Jiaozuo Henan Peoples R China

ISBN: (纸本)9798350307443

Deep learning for unsupervised image segmentation remains challenging due to the absence of human labels. The common idea is to train a segmentation head, with the supervision of pixel-wise pseudo-labels generated based on the representation of self-supervised backbones. By doing so, the model performance depends much on the distance between the distribution of target datasets, and the one of backbones' pre-training dataset (e.g., imageNet). In this work, we investigate a new task, namely unsupervised camouflaged object segmentation (UCOS), where the target objects own a common rarely-seen attribute, i.e., camouflage. Unsurprisingly, we find that the state-of-the-art unsupervised models struggle in adapting UCOS, due to the domain gap between the properties of generic and camouflaged objects. To this end, we formulate the UCOS as a source-free unsupervised domain adaptation task (UCOS-DA), where both source labels and target labels are absent during the whole model training process. Specifically, we define a source model consisting of self-supervised vision transformers pre-trained on imageNet. On the other hand, the target domain includes a simple linear layer (i.e., our target model) and unlabeled camouflaged objects. We then design a pipeline for foreground-background-contrastive selfadversarial domain adaptation, to achieve robust UCOS. As a result, our baseline model achieves superior segmentation performance when compared with competing unsupervised models on the UCOS benchmark, with the training set which's scale is only one tenth of the supervised COS counterpart. The UCOS benchmark and our baseline model are now publicly available(1).

关键词： Source Free Domain Adaptation unsupervised image segmentation

来源：评论

学校读者我要写书评

暂无评论

unsupervised fuzzy clustering using Weighted Incremental Neural Networks.

引用

International journal of neural systems 2004年第6期14卷 355-371页

作者： Muhammed, Hamed Hamid Centre for Image Analysis Uppsala University Lagerhyddsvagen 3 Uppsala SE-75237 Sweden

A new more efficient variant of a recently developed algorithm for unsupervised fuzzy clustering is introduced. A Weighted Incremental Neural Network (WINN) is introduced and used for this purpose. The new approach is called FC-WINN (Fuzzy Clustering using WINN). The WINN algorithm produces a net of nodes connected by edges, which reflects and preserves the topology of the input data set. Additional weights, which are proportional to the local densities in input space, are associated with the resulting nodes and edges to store useful information about the topological relations in the given input data set. A fuzziness factor, proportional to the connectedness of the net, is introduced in the system. A watershed-like procedure is used to cluster the resulting net. The number of the resulting clusters is determined by this procedure. Only two parameters must be chosen by the user for the FC-WINN algorithm to determine the resolution and the connectedness of the net. Other parameters that must be specified are those which are necessary for the used incremental neural network, which is a modified version of the Growing Neural Gas algorithm (GNG). The FC-WINN algorithm is computationally efficient when compared to other approaches for clustering large high-dimensional data sets.

关键词： data reduction Growing Neural Gas (GNG) incremental neural networks neuro-fuzzy systems unsupervised fuzzy clustering unsupervised image segmentation watersheds

来源：评论

学校读者我要写书评

暂无评论

A Comparative Study on Lagrange Ying-Yang Alternation Method in Gaussian Mixture-Based Clustering 18th

A Comparative Study on Lagrange Ying-Yang Alternation Method...

引用

18th International Conference on Intelligent Data Engineering and Automated Learning (IDEAL)

作者： Long, Weijian Tu, Shikui Xu, Lei Shanghai Jiao Tong Univ Dept Comp Sci & Engn Shanghai Peoples R China Shanghai Jiao Tong Univ Ctr Cognit Machines & Computat Hlth Shanghai Peoples R China Chinese Univ Hong Kong Dept Comp Sci & Engn Hong Kong Hong Kong Peoples R China

ISBN: (纸本)9783319689357;9783319689340

Gaussian Mixture Model (GMM) has been applied to clustering with wide applications in image segmentation, object detection and so on. Many algorithms were proposed to learn GMM with appropriate number of Gaussian components automatically determined. Lagrange Ying-Yang alternation method (LYYA) is one of them and it has advantages of no priors as well as the posterior probability bounded by traditional probability space. This paper aims to investigate the performance of LYYA, in comparisons with other methods including Bayesian Ying-Yang (BYY) learning, Rival penalized competitive learning (RPCL), hard-cut Expectation Maximization (EM) method, and classic EM with Bayesian Information Criterion (BIC). Systematic simulations show that LYYA is generally more robust than others on the data generated by varying sample size, data dimensionality and real components number. unsupervised image segmentation results on Berkeley datasets also confirm LYYA advantages when comparing to the Mean shift and Multiscale graph decomposition algorithms.

关键词： Gaussian Mixture Model Lagrange Ying-Yang alternation method unsupervised image segmentation Lagrange coefficient

来源：评论

学校读者我要写书评

暂无评论

Texture-based segmentation of high resolution SAR images using contourlet transform and mean shift

Texture-based segmentation of high resolution SAR images usi...

引用

IEEE International Conference on Information Acquisition

作者： Li Yingqi He Mingyi Northwestern Polytech Univ Coll Elect Engn 127 W Youyi Rd Xian 710072 Shaanxi Peoples R China

ISBN: (纸本)1424405289

This paper presents an unsupervised texture-based segmentation algorithm which uses reduced contourlet transform sub-bands and mean shift clustering, to analysis the texture information of high resolution SAR images. One step and criteria is proposed to reduce the sub-bands and other's is presented to decrease the number of dimension of the feature space. The mean shift clustering method is used to obtain the number of texture regions and the centre of the label class. Group the pixels into corresponding texture region by their simple distance to the class centre pixel. Experiments on a mixture of Brodatz texture and SAR images show the proposed algorithm of using contourlet transform and mean shift clustering gives satisfactory results.

关键词： SAR texture unsupervised image segmentation contourlet transform mean shift feature selection

来源：评论

学校读者我要写书评

暂无评论

Increasing the Accuracy and Automation of Fractional Vegetation Cover Estimation from Digital Photographs

引用

REMOTE SENSING 2016年第7期8卷 474-474页

作者： Coy, Andre Rankine, Dale Taylor, Michael Nielsen, David C. Cohen, Jane Univ West Indies Dept Phys Mona Jamaica USDA ARS Cent Great Plains Res Stn Akron CO 80720 USA Univ West Indies Dept Life Sci Mona Jamaica

The use of automated methods to estimate fractional vegetation cover (FVC) from digital photographs has increased in recent years given its potential to produce accurate, fast and inexpensive FVC measurements. Wide acceptance has been delayed because of the limitations in accuracy, speed, automation and generalization of these methods. This work introduces a novel technique, the Automated Canopy Estimator (ACE) that overcomes many of these challenges to produce accurate estimates of fractional vegetation cover using an unsupervised segmentation process. ACE is shown to outperform nine other segmentation algorithms, consisting of both threshold-based and machine learning approaches, in the segmentation of photographs of four different crops (oat, corn, rapeseed and flax) with an overall accuracy of 89.6%. ACE is similarly accurate (88.7%) when applied to remotely sensed corn, producing FVC estimates that are strongly correlated with ground truth values.

关键词： fractional vegetation cover automated canopy estimation unsupervised image segmentation digital photographs

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：