检索结果-内蒙古大学图书馆

23rd IEEE international conference on image processing (ICIP)

作者： Yamaguchi, Takuro Ikehara, Masaaki Keio Univ EEE Dept Yokohama Kanagawa 2238522 Japan

ISBN: (纸本)9781467399616

image upsampling from one input image gathers considerable attention in the field of computer vision. The problem is ill-posed because the number of known low-resolution (LR) pixels is less than that of unknown high-resolution (HR) pixels. Therefore, quality of an up sampled image depends on prior assumptions. image interpolation methods are one of the image upsampling technologies and are faster than other image upsampling technologies such as Super-Resolution. However, these methods tend to cause jaggies and blurs in edge and texture regions. We use the idea of Multi-surface Fitting (MF) to solve these problems. MF uses plural local functions to estimate an HR pixel and it reduces blurs. Moreover, we utilize filtering instead of calculation of each local function in order to reduce a computational cost. And we introduce new weights to estimate edge directions. By these ideas, our method has both high quality and a low computational cost.

关键词： image interpolation filtering multi-surface fitting weighted mean

来源：评论

学校读者我要写书评

暂无评论

vision-based surface roughness evaluation system for end milling

引用

international JOURNAL OF computer INTEGRATED MANUFACTURING 2018年第8期31卷 727-738页

作者： Cuka, Besmir Cho, Minho Kim, Dong-Won Chonbuk Natl Univ Ind & Informat Syst Engn Jeonju South Korea Korea Marine Equipment Res Inst Gunsan Si Jeollabuk Do South Korea

This paper presents a vision-based surface roughness evaluation system for end-milled metals, addressing digital reconstruction and calibration of inspected surfaces, and quantitative and qualitative evaluation of surface texture. Specimens with different levels of surface roughness are machined, and a comparison between stylus-based and vision-based measurements is performed while using standard roughness parameters. The vison-based results vary among 9% and 11% compared to the stylus-based ones, which is a minor error to trade-off for faster measurements. Furthermore, surface texture evaluation is performed by detecting the generated cusp lines and tool marks on the machined surface. The tool marks' distribution is analysed in order to determine whether the machining is performed under optimal cutting conditions. Results show that under optimal cutting conditions, the detected tool marks are normally distributed along the feed direction and the distance between two consecutive tool marks does not vary significantly. Based on the proposed methods software is implemented that enables the three-dimensional reconstruction, calibration and evaluation of the inspected surface.

关键词： Surface roughness computer vision image processing end milling

来源：评论

学校读者我要写书评

暂无评论

Algorithms and Approaches for processing an Output Signal from Lab-on-a-Chip Devices in Bioanalysis 12

Algorithms and Approaches for Processing an Output Signal fr...

引用

12th international conference on Perspective Technologies and Methods in MEMS Design (MEMSTECH)

作者： Lizanets, Danylo Walczak, Rafal Wroclaw Univ Technol Fac Microelect & Photon Div Microengn & Photovolta PL-50370 Wroclaw Poland

ISBN: (纸本)9786176079132

Algorithms and approaches used for implementing software for processing an image output signal from microfluidic devices are presented in this paper. Such microdevices (often called a lab-on-a-chip) are utilized in analysis of biological cells;their behaviors are investigated under different sets of conditions. We present here algorithms for cell detecting, processing and automatic recognition of shape and size of cells in lab-chips for bioanalysis.

关键词： image processing lab-on-a-chip computer vision bioanalysis

来源：评论

学校读者我要写书评

暂无评论

AE-CNN Based Supervised image Classification 5th

AE-CNN Based Supervised Image Classification

引用

5th international conference on computer vision and image processing, cvip 2020

作者： Chandra, Ganduri Challa, Muralidhar Reddy Indian Institute of Technology Hyderabad Hyderabad India

ISBN: (纸本)9789811611025

Point of Care Ultrasound (PoCUS) imaging is an important tool in detecting lung consolidations and tissue sliding, and hence has a potential to identify the onset of novel-CoVID-19 attack in a person. Of late, Convolutional Neural Network (CNN) architectures have gained popularity in improving the accuracy of the predictions. Motivated by this, in this paper, we introduce a CNN based Auto Encoder (AE-CNN) for a better representation of the features to get an accurate prediction. While most of the existing models contain ‘fully connected’ (FC) layers, in our work, we use only convolutional layers instead of FC layers before the output layer, which helps us in achieving a less training time of the model. Moreover, fully connected layers of a network can not learn the patterns in an image as much as convolutional layers can. This is the main advantage of our model over its existing counterparts. We demonstrate that our model detects the lung abnormalities in the ultrasound images with an accuracy of 96.6%. © 2021, Springer Nature Singapore Pte Ltd.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Brain MRI and CT image Fusion Using Generative Adversarial Network 1

引用

6th international conference on computer vision and image processing, cvip 2021

作者： Narute, Bharati Bartakke, Prashant College of Engineering MH Pune411001 India

ISBN: (数字)9783031113499

ISBN: (纸本)9783031113482

The generative adversarial networks (GAN), complete model, is used to fuse computed tomography (CT) and magnetic resonance imaging (MRI) brain images in this research paper. To create a resultant fused image with bone structures from CT images and soft tissues from MRI images, our method develops an adversarial game between a generator and a discriminator. To make a stable training process, we use GAN instead of conventional fusion methods, and our architecture can handle different resolutions of multi-source medical images. The efficacy of the proposed procedure is demonstrated using several evaluation metrics. The proposed algorithms provide the best fused images without distortion and false artefacts. Comparison of proposed methods is done with the conventional techniques. The images obtained by fusing both sources’ content with the help of the above algorithm gives the best with respect to visualization and diagnosis of the condition. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Enhancing stereo image quality based on adaptive hole-filling of depth image for kinect camera

Enhancing stereo image quality based on adaptive hole-fillin...

引用

2016 international conference on image processing, computer vision, and Pattern Recognition, IPCV 2016

作者： Cho, Ji-Min Yun, Young-Ji Nam, Seung-Woo Chien, Sung-Il School of Electronics Engineering Kyungpook National University Daegu Korea Republic of SW Content Research Lab Electronics and Telecommunications Research Institute Daejeon Korea Republic of

ISBN: (纸本)1601324421

Kinect camera produces depth images as well as color images. Yet undefined depth regions called holes are often observed in the depth image and are quite harmful in generating a second view image to construct a stereo pair for 3D viewing. An adaptive hole-filling method that can fill in even very large holes more naturally is proposed so that depth values for holes of various sizes are suitably estimated depending on the location of holes in the background or foreground. Removal of dot noise inside a hole is also introduced to further enhance the quality of the newly generated stereo image. © CSREA Press.

关键词： Filling

来源：评论

学校读者我要写书评

暂无评论

Single image Dehazing Using Multipath Networks Based on Chain of U-Nets 1

引用

7th international conference on computer vision and image processing, cvip 2022

作者： Deivalakshmi, S. Sandana, J. Sudaroli Department of Electronics and Communication Engineering National Institute of Technology Tiruchirappalli India

ISBN: (数字)9783031314070

ISBN: (纸本)9783031314063

Single-image haze removal is a necessary pre-processing step for several computer vision applications and may be a challenging task. The images captured outdoors are seriously degraded in color and contrast due to extreme weather conditions like fog, haze, snow and rain. This poor visual quality can inhibit the performance, which is intended to operate on clear conditions like object detection. In recent years, single-image dehazing, which recovers clean images from haze-affected images, has gained popularity. This paper proposes a unique and efficient multipath supported U-Net called Ladder-Net to directly restore a hazy free image from a hazy image. Utilizing only a single set of encoder and decoder branches within the U-Net structure, the proposed Ladder-Net has numerous encoder and decoder branches and dense skip connections between each pair of adjoining encoder and decoder branches at every level. The encoder and decoder are designed to acquire the contextual information in the input images and to evaluate the role of each input to the resulting hazy-free images, respectively, without relying on the atmospheric scattering model. The proposed model is trained and tested using RESIDE benchmark dataset. Experiments conducted on hazy images and a synthetic dataset obtained naturally show that the proposed method proves effective in achieving superior performance over previous methods. The experimental results convey that the proposed architecture achieves exceptional performance over the prevailing methods concerning the qualitative and the quantitative performance, PSNR and SSIM. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Human activity recognition by fusion of RGB, depth, and skeletal data 2nd

Human activity recognition by fusion of RGB, depth, and skel...

引用

2nd international conference on computer vision and image processing, cvip 2017

作者： Khaire, Pushpajit Imran, Javed Kumar, Praveen Department of Computer Science and Engineering Visvesvaraya National Institute of Technology Nagpur India Department of Computer Science and Engineering Indian Institute of Technology Roorkee India

ISBN: (纸本)9789811078941

A significant increase in research of human activity recognition can be seen in recent years due to availability of low-cost RGB-D sensors and advancement of deep learning algorithms. In this paper, we augmented our previous work on human activity recognition (Imran et al., IEEE international conference on advances in computing, communications, and informatics (ICACCI), 2016) [1] by incorporating skeletal data for fusion. Three main approaches are used to fuse skeletal data with RGB, depth data, and the results are compared with each other. A challenging UTD-MHAD activity recognition dataset with intraclass variations, comprising of twenty-seven activities, is used for testing and experimentation. Proposed fusion results in accuracy of 95.38% (nearly 4% improvement over previous work), and it also justifies the fact that recognition improves with an increase in number of evidences in support. © 2018, Springer Nature Singapore Pte Ltd.

关键词： Musculoskeletal system

来源：评论

学校读者我要写书评

暂无评论

An automated ear localization technique based on modified hausdorff distance

An automated ear localization technique based on modified ha...

引用

international conference on computer vision and image processing, cvip 2016

作者： Sarangi, Partha Pratim Panda, Madhumita Mishra, B.S.P. Dehuri, Sachidananda School of Computer Engineering KIIT University Bhubaneswar India Department of MCA Seemanta Engineering College Jharpokharia India Department of ICT FM University Balasore India

ISBN: (纸本)9789811021060

Localization of ear in the side face images is a fundamental step in the development of ear recognition based biometric systems. In this paper, a well-known distance measure termed as modified Hausdorff distance (MHD) is proposed for automatic ear *** introduced theMHD to decrease the effect of outliers and allowing it more suitable for detection of ear in the side face images. The MHD uses coordinate pairs of edge pixels derived from ear template and skin regions of the side face image to locate the ear portion. To detect ears of various shapes, ear template is created by considering different structure of ears and resized it automatically for the probe image to find exact location of ear. The CVL and UND-E database have side face images with different poses, inconsistent background and poor illumination utilized to analyse the effectiveness of the proposed algorithm. Experimental results reveal the strength of the proposed technique is invariant to various poses, shape, occlusion, and noise. © Springer Science+Business Media Singapore 2017.

关键词： Biometrics

来源：评论

学校读者我要写书评

暂无评论

Unsupervised image to image Translation for Multiple Retinal Pathology Synthesis in Optical Coherence Tomography Scans 7th

Unsupervised Image to Image Translation for Multiple Retinal...

引用

7th international conference on computer vision and image processing, cvip 2022

作者： Pasupuleti, Hemanth Kothari, Abhishek R. Girish, G.N. Computer Science and Engineering Group Indian Institute of Information Technology Sri City India Pink City Eye and Retina Center Jaipur India

ISBN: (纸本)9783031314162

image to image Translation (I2I) is a challenging computer vision problem used in numerous domains for multiple tasks. Recently, ophthalmology became one of the major fields where the application of I2I is increasing rapidly. One such application is the generation of synthetic retinal optical coherence tomographic (OCT) scans. Existing I2I methods require training of multiple models to translate images from normal scans to a specific pathology: limiting the use of these models due to their complexity. To address this issue, we propose an unsupervised multi-domain I2I network with pre-trained style encoder that translates retinal OCT images in one domain to multiple domains. We assume that the image splits into domain-invariant content and domain-specific style codes, and pre-train these style codes. The performed experiments show that the proposed model outperforms state-of-the-art models like MUNIT and CycleGAN by synthesizing diverse pathological scans with lower FID scores (108.30-Kermany dataset;60-TaeKeun datset) and higher LPIPS scores (0.146-Kermany dataset;0.171-TaeKeun dataset). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

关键词： Generative adversarial networks

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：