检索结果-内蒙古大学图书馆

IEEE International conference on Consumer Electronics - Taiwan (ICCE-TW)

作者： Sintunata, Vicky Aoki, Terumasa Tohoku Univ Grad Sch Informat Sci Sendai Miyagi Japan Tohoku Univ New Ind Hatchery Ctr Sendai Miyagi Japan

ISBN: (纸本)9781509040179

Conventional skeleton extraction methods require a closed boundary constraint to solve the problem. In natural images closed boundary constraint might not be easily satisfied due to the similarity of the object and its background, occlusion, etc. In this paper a novel approach based on the Delaunay triangulation to solve the skeleton extraction in natural images is proposed. The algorithm shows a promising result in extracting the skeleton in natural images.

关键词： Skeleton image edge detection image color analysis Clustering algorithms Transforms Gray-scale Standards

来源：评论

学校读者我要写书评

暂无评论

Tukey's Biweight Constrain Function For A Surrogate image Enlargement Based On SISR Algorithm Using High Spectrum Estimation 8

Tukey's Biweight Constrain Function For A Surrogate Image En...

引用

8th International conference of Information and Communication Technology for Embedded systems (IC-ICTES)

作者： Thakulsukanant, Kornkamol Patanavijit, Vorapoj Assumption Univ Bangkok Thailand

ISBN: (纸本)9781509048090

Consistently, the classical image enlargement algorithm is a scientific analytical method for producing a better refined resolution image that is frequently required for advanced digital image processing (DIP) from a single lower resolution image that is frequently acquired from digital camera embedded system. Because of its less computation calculation, the Single-image Super-Resolution (SISR) that analytically applied for a single lower resolution image is one of the worldwide effective Super Resolution-Reconstruction (SRR) algorithms thus this paper proposes the image enlargement based on the SISR algorithm using high spectrum estimation and Tukey's Biweight constrain function. In general, the performance of this SISR algorithm is hinge on up to three parameters (b, h, k) however there are burdensome for determining these optimized values for these parameters (b, h, k). In order to solve this problem, the Tukey's Biweight constrain function, which is hinge on merely single parameter (T), contrary to three parameters like the classical constrain function, is engaged in the SISR algorithm. By examining on 14 benchmark images, which are profaned by considerable noise forms, in scientific analytical scrutinizing sector, the novel SISR algorithm illustrates that there is efficiently and effortlessly in parameter setting process but the performance of the novel SISR algorithm (with single parameters) is nearly equal to the original SISR (with three parameters). Due to greatly time reduction in the parameter setting process, this novel SISR algorithm is more advisable for real-time applications.

关键词： Tukey's Biweight SISR (Single-image Super-Resolution) SRR (Super Resolution-Reconstruction) Digital image Enlargement

来源：评论

学校读者我要写书评

暂无评论

Joint autoregressive and hierarchical priors for learned image compression 18

Joint autoregressive and hierarchical priors for learned ima...

引用

Proceedings of the 32nd International conference on Neural Information processing systems

作者： David Minnen Johannes Ballé George Toderici Google Research

Recent models for learned image compression are based on autoencoders that learn approximately invertible mappings from pixels to a quantized latent representation. The transforms are combined with an entropy model, which is a prior on the latent representation that can be used with standard arithmetic coding algorithms to generate a compressed bitstream. Recently, hierarchical entropy models were introduced as a way to exploit more structure in the latents than previous fully factorized priors, improving compression performance while maintaining end-to-end optimization. Inspired by the success of autoregressive priors in probabilistic generative models, we examine autoregressive, hierarchical, and combined priors as alternatives, weighing their costs and benefits in the context of image compression. While it is well known that autoregressive models can incur a significant computational penalty, we find that in terms of compression performance, autoregressive and hierarchical priors are complementary and can be combined to exploit the probabilistic structure in the latents better than all previous learned models. The combined model yields state-of-the-art rate–distortion performance and generates smaller files than existing methods: 15.8% rate reductions over the baseline hierarchical model and 59.8%, 35%, and 8.4% savings over JPEG, JPEG2000, and BPG, respectively. To the best of our knowledge, our model is the first learning-based method to outperform the top standard image codec (BPG) on both the PSNR and MS-SSIM distortion metrics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Real-Time Lane Detection Using Spatio-Temporal Incremental Clustering 20

Real-Time Lane Detection Using Spatio-Temporal Incremental C...

引用

20th IEEE International conference on Intelligent Transportation systems (ITSC)

作者： Gupta, Any Choudhary, Ayesha Jawaharlal Nehru Univ Sch Comp & Syst Sci New Delhi India

ISBN: (纸本)9781538615263

In this paper, we propose a novel method for lane detection in real-time based on unsupervised learning using images from a camera mounted on the dashboard of a vehicle. Lane detection is a crucial element of an intelligent vehicle safety system. Lane departure warning systems rely on accurate detection of lanes and are important for a driver assistant system. In our system, we extract the features of interest and detect the lane markings or foreground regions in each frame. We develop a spatio-temporal incremental clustering algorithm coupled with curve fitting for detecting the lanes on-the-fly. The spatio-temporal incremental clustering is performed over the detected foreground region for finding the lanes accurately. Each cluster represents a lane across space and time. The lanes are then identified by fitting curves on these clusters. Our system is capable of accurately detecting straight and curved lanes, noncontinuous and continuous lanes and is independent of number of lanes in the frame and the system is fast because we have not used any database for processing the images. Experimental results show that our algorithm can accurately perform lane detection in challenging scenarios.

关键词： Roads Clustering algorithms Cameras Real-time systems Feature extraction image color analysis Unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Azimuthally invariant Mueller-matrix mapping of biological optically anisotropic network 40

Azimuthally invariant Mueller-matrix mapping of biological o...

引用

conference on Applications of Digital image processing XL

作者： Ushenko, Yu. O. Vanchuliak, O. Bodnar, G. B. Ushenko, V. O. Grytsyuk, M. Pavlyukovich, N. Pavlyukovich, O. V. Antonyuk, O. Chernivtsi Natl Univ 2 Kotsyubinsky Str UA-58012 Chernovtsy Ukraine Bukovinian State Med Univ UA-58000 Chernovtsy Ukraine

ISBN: (数字)9781510612501

ISBN: (纸本)9781510612501;9781510612495

A new technique of Mueller-matrix mapping of polycrystalline structure of histological sections of biological tissues is suggested. The algorithms of reconstruction of distribution of parameters of linear and circular dichroism of histological sections liver tissue of mice with different degrees of severity of diabetes are found. The interconnections between such distributions and parameters of linear and circular dichroism of liver of mice tissue histological sections are defined. The comparative investigations of coordinate distributions of parameters of amplitude anisotropy formed by Liver tissue with varying severity of diabetes (10 days and 24 days) are performed. The values and ranges of change of the statistical (moments of the 1st - 4th order) parameters of coordinate distributions of the value of linear and circular dichroism are defined. The objective criteria of cause of the degree of severity of the diabetes differentiation are determined.

关键词： polarization Fourier optics and signal processing imaging systems medical and biological imaging

来源：评论

学校读者我要写书评

暂无评论

Cartooning for Enhanced Privacy in Lifelogging and Streaming Videos 30

Cartooning for Enhanced Privacy in Lifelogging and Streaming...

引用

30th IEEE/CVF conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

作者： Hassan, Eman T. Hasan, Rakibul Shaffer, Patrick Crandall, David Kapadia, Apu Indiana Univ Bloomington Sch Informat & Comp Bloomington IN 47405 USA

ISBN: (纸本)9781538607336

We describe an object replacement approach whereby privacy-sensitive objects in videos are replaced by abstract cartoons taken from clip art. Our approach uses a combination of computer vision, deep learning, and image processing techniques to detect objects, abstract details, and replace them with cartoon clip art. We conducted a user study (N=85) to discern the utility and effectiveness of our cartoon replacement technique. The results suggest that our object replacement approach preserves a video's semantic content while improving its privacy by obscuring details of objects.

关键词： Estimation Field programmable gate arrays OFDM Modulation Filtration Clocks Signal processing algorithms

来源：评论

学校读者我要写书评

暂无评论

Playback detection using machine learning with spectrogram features approach 10

Playback detection using machine learning with spectrogram f...

引用

10th International conference on Human System Interactions (HSI)

作者： Dembski, Jerzy Ruminski, Jacek Gdansk Univ Technol Fac Elect Telecommun & Informat Ul Narutowicza 11-12 PL-80952 Gdansk Poland

ISBN: (纸本)9781509046881

This paper presents 2D image processing approach to playback detection in automatic speaker verification ( ASV) systems using spectrograms as speech signal representation. Three feature extraction and classification methods: histograms of oriented gradients ( HOG) with support vector machines ( SVM), HAAR wavelets with AdaBoost classifier and deep convolutional neural networks ( CNN) were compared on different data partitions in respect of speakers or playback devices: for instance with different speakers in training and test subsets. The playback detection systems were trained and tested on two speech datasets S-1 and S-2 manufactured independently by two different institutions. The test error for both datasets oscillates about the level of 1% for HOG+SVM and even below it for CNN in bigger S-1 base. In cross validation scenario in which one base was used for training and second base for the test the results were very poor what suggests that the information relevant for playback detection appeared in each base in different way.

关键词： Estimation Field programmable gate arrays OFDM Modulation Filtration Clocks Signal processing algorithms

来源：评论

学校读者我要写书评

暂无评论

Fast algorithms and Architectures for 8-Point DST-II/DST-vii Approximations

引用

JOURNAL OF CIRCUITS systems AND COMPUTERS 2017年第3期26卷 1750045-1750045页

作者： Cintra, Renato J. Bayer, Fabio M. Madanayake, Arjuna Potluri, Uma S. Edirisuriya, Amila Univ Fed Pernambuco Signal Proc Grp Dept Stat Recife PE Brazil Inst Natl Sci Appl LIRIS Lyon France Univ Rennes 1 INRIA IRISA Equipe Cairn Rennes France Univ Fed Santa Maria Dept Stat Santa Maria RS Brazil Univ Fed Santa Maria LACESM Santa Maria RS Brazil Univ Akron ECE Akron OH 44325 USA

Multiplier-free fast algorithms are derived and analyzed for realizing the 8-point discrete sine transform of type II and type vii (DST-II and DST-vii) transforms with applications in image and video compression. A new fast algorithm is identified using numerical search methods for approximating DST-vii without employing multipliers. In addition, recently proposed fast algorithms for approximating the 8-point DCT-II are now extended to approximate DST-II. All proposed approximations for DST-II and DST-vii are compared with ideal transforms, and circuit complexity is measured using FPGA-based rapid prototypes on a 90nm Xilinx Virtex-4 device. The proposed architectures find applications in emerging video processing standards such as H.265/HEVC.

关键词： Approximate discrete transforms DCT DST fast algorithms FPGA low-complexity transform

来源：评论

学校读者我要写书评

暂无评论

Leveraging Deep Preference Learning for Indexing and Retrieval of Biomedical images 8

Leveraging Deep Preference Learning for Indexing and Retriev...

引用

8th International IEEE/EMBS conference on Neural Engineering (NER)

作者： Pang, Shuchao Orgun, Mehmet A. Du, Anan Yu, Zhezhou Jilin Univ Coll Comp Sci & Technol Qianjin St 2699 Changchun Jilin Peoples R China Macquarie Univ Dept Comp Sydney NSW 2109 Australia China Mobile HangZhou Informat Technol Co Ltd Hangzhou Zhejiang Peoples R China

ISBN: (纸本)9781509046034

This paper presents an original framework based on deep learning and preference learning to retrieve and characterize biomedical images for assisting physicians in diagnosing complex diseases with potentially only small differences between them. In particular, we use deep learning to extract the high-level and compact features for biomedical images. In contrast to the traditional biomedical algorithms or general image retrieval systems that only consider the use of pixel and/or hand-crafted features to represent images, we utilize deep neural networks for feature discovery of biomedical images. Moreover, in order to be able to index the similarly referenced images, we introduce preference learning in a novel way to learn what kinds of images we need so that we can obtain the similarity ranking list of biomedical images. We evaluate the performance of our system in detailed experiments over the well-known available OASIS-MRI database for whole brain neuroimaging as a benchmark and compare it with those of the traditional biomedical and general image retrieval approaches. Our proposed system exhibits an outstanding retrieval ability and efficiency for biomedical image applications.

关键词： Estimation Field programmable gate arrays OFDM Modulation Filtration Clocks Signal processing algorithms

来源：评论

学校读者我要写书评

暂无评论

A Novel Algorithm for Laser Spot Detection Based on Quaternion Discrete Cosine Transform

A Novel Algorithm for Laser Spot Detection Based on Quaterni...

引用

International conference on Intelligent and Interactive systems and Applications (IISA)

作者： Miao, Xiaodong Shi, Xiang Xu, Zhixing Wang, Aiqing Wang, Wei Kangni Mech & Elect Co Ltd Nanjing 210009 Jiangsu Peoples R China Nanjing Univ Technol Sch Mech & Power Engn Nanjing 211800 Jiangsu Peoples R China

ISBN: (纸本)9783319495682;9783319495675

Laser spot detection is an important problem in optical measurement. The precision and speed of the detection algorithm influence the optical measurement system directly. The traditional algorithms such as Hough Transform and Gravity model are unsatisfactory in complex conditions. The laser spot detection algorithm referred in this paper is based on the quaternion discrete cosine transform and moment and a method is adopted to approximate the edge of the laser spot. Not only the center and edge can be detected simultaneously but also the robustness of noise is better than others. The algorithm is suitable for the real-time optical measurement.

关键词： Laser spot image processing Quaternion discrete cosine transform

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：