检索结果-内蒙古大学图书馆

Application of superpixel segmentation and morphological projector for structural changes detection in remote sensing images

引用

Journal of Physics: conference Series 2019年第3期1368卷

作者： A Y Denisova A A Egorova V V Sergeyev Samara National Research University Moskovskoe Shosse 34A Samara Russia 443086 Image Processing Systems Institute of RAS - Branch of the FSRC 'Crystallography and Photonics' RAS Molodogvardejskaya street 151 Samara Russia 443001

Detection of structural changes in images is one of the important tasks of remote sensing (RS) data thematic analysis. The effective way to solve it is applying the Pyt'ev's morphological projector to the pair of images of the same scene acquired on different dates. The main advantage of this method is its invariance to global brightness transformations, which in the case of RS images correspond to different parameters of the atmosphere or the different values of the brightness-contrast ratio of the scene. However, the classical Pyt'ev's morphological projector and its regularized versions do not take into account the spatial connectivity of image samples. As a result, they ignore the textural features of images. In this article, we suggest the algorithm of structural changes detection based on superpixel segmentation and Pyt'ev's morphological projector that takes into account local characteristics of the image pixels. In the experimental research, we analyzed the accuracies of the proposed and classical Pyt'ev's structural change detection methods using simulated and real RS images. The comparison of two algorithms showed that the proposed method is more robust to the additive white Gaussian noise (AWGN) at different values of signal-to-noise (SNR) ratio. Additionally, the experiments with nonlinear brightness distortions (vignetting) of one of the pair of images demonstrated that the proposed method has lower false positive rates than the classical one.

关键词：

来源：评论

学校读者我要写书评

暂无评论

The system for determining the human head position and orientation for vehicle simulators 13

The system for determining the human head position and orien...

引用

13th International Scientific conference on Optoelectronic Equipment and Devices in systems of Pattern Recognition, image and Symbol Information processing, RECOGN 2017

作者： Zakharov, Alexei Barinov, Alexei Zhiznyakov, Arkady Murom Institute Vladimir State University Murom Russia

The system for determining the human head position and orientation for vehicle simulators is considered in the work. It is necessary to take into account the current position and orientation of the driver's head when simulating a virtual space. The model for determining the human head position and orientation is developed and studied. In the course of the study, the developed position and orientation algorithm was compared with analogs belonging. This model is characterized by the joint use of three-dimensional reconstruction, stereo vision, the spectral theory of graphs, spectral embedding of graphs into a vector subspace, and allowing the head to rotate to 50° with an accuracy of 3°, which exceeds known approaches. A system for determining the human head position and orientation based on stereo images is constructed, which implements the developed algorithms. The developed system can be represented in the form of the following structure: stereo module, initialization module, tracking module, module for calculating the angles of the head orientation, head detection module, information transfer module.

关键词： image recognition

来源：评论

学校读者我要写书评

暂无评论

Random Sparse Representation for Thermal to Visible Face Recognition

Random Sparse Representation for Thermal to Visible Face Rec...

引用

IEEE Symposium on Computers and Communications (ISCC)

作者： Reihanian, Samira Arbabi, Ehsan Maham, Behrouz Univ Tehran Sch ECE Tehran Iran Nazarbayev Univ Sch Engn Astana Kazakhstan

ISBN: (纸本)9781538616291

Heterogeneous face recognition (HFR) has a prominent importance in sophisticated face recognition systems. Thermal to visible scenario, where the gallery and the probe images are respectively captured in visible and long wavelength infrared (LWIR) band, is one of the most challenging and interesting HFR scenarios. Since the formation of thermal images does not require an external illumination source, the deployment of thermal probe images is practical even in totally darkness conditions such as night security surveillance systems. In this paper, we propose an ensemble classifier which uses the random subspace idea for defining different representations of each image in distinct base learners, and exploits the sparse representation algorithm for the classification of thermal probe images. According to the experimental results, our proposed algorithm leads significant performance improvements in the area of thermal to visible face recognition and achieves the average Rank-1 accuracy of 89.33 percent.

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

Multi process matching depth image between stereo camera and structure sensor in realtime 18

Multi process matching depth image between stereo camera and...

引用

18th International Electronics Symposium on Engineering Technology and Applications, IES-ETA 2017

作者： Anjarlistiawan, Rimsya Farouq, Ardiansyah Al Sukaridhoto, Sritrusta Dewanto, Raden Sanggar Pramadihanto, Dadet Surabaya Indonesia Surabaya Indonesia

ISBN: (纸本)9781538607121

A map of a location made of a human mindset is made briefly by recording a situation already seen in memory. Human behavior is easy to do without a lot of thinking. The process is observed more deeply that when the human mind does not do the behavior in turn like looking at the whole room then just remember, but humans can see at once remember simultaneously. Bringing human behavior into a computer system is done by mimicking how humans look. Looking at using two cameras in combination with the sensor structure. To process the device requires a large computing. Be a challenge for Portable CPU that have lower capabilities than high end CPU. For that humanoid robot can perform tasks simultaneously and can accelerate, especially in image processing using portable CPU required a method that makes the processing into realtime. Achieving realtime with data from previous researcher can not be done with sequential algorithms when the complexity is high. After analyzing the problem was built parallel camera mapping process using multithread algorithm. This paper aims to accelerate the basic process of making the map so as to achieve realtime. This research aims to change the previous process that runs sequentially into a parallel system using multithread algorithm. The results of each process are compared between parallel systems with multithread algorithm and do not use parallel systems. There is a change in speed between the two processes. Further research is concerned with improving the mapping environment results when visualized. © 2017 IEEE.

关键词： Mapping

来源：评论

学校读者我要写书评

暂无评论

image Analysis using Convolutional Neural Networks for Modeling 2D Fracture Propagation 17

Image Analysis using Convolutional Neural Networks for Model...

引用

17th IEEE International conference on Data Mining (ICDMW)

作者： Miller, Robyn L. Moore, Bryan Viswanathan, Hari Srinivasan, Gowri Los Alamos Natl Lab Los Alamos NM 87545 USA

ISBN: (纸本)9781538638002

The primary failure mechanism in brittle materials such as ceramics, granite and some metal alloys is through the presence of defects which result in crack formation and propagation under the application of load. We are interested in studying this process of crack propagation, interaction and coalescence, which degrades the strength of the specimen. Traditionally, engineering applications that study these materials employ finite element mesh-based methods that require hundreds of hours of processing time on multi-core high performance clusters. We have developed a graph-based reduced order model that captures key geometric and topological features of the dynamic fracture propagation network. We report here the early stages of our study in which deep neural networks will be applied to dynamic directed weighted graphs capturing various metrics of crack-pair interaction strength with the aim of predicting crack lengths, dynamic crack growth/coalescence properties, distributions of these properties over the entire material through time, failure paths and time to failure. Our graph-based representations allow us to consider detailed topology in conjunction with metric geometry to gain insights into the dominant mechanisms that drive the physics in these systems.

关键词： Physics Feature extraction Data models Integrated circuits Data mining Geometry Heuristic algorithms

来源：评论

学校读者我要写书评

暂无评论

Hair and Bare Skin Discrimination for Laser-Assisted Hair Removal systems 39

Hair and Bare Skin Discrimination for Laser-Assisted Hair Re...

引用

39th Annual International conference of the IEEE-Engineering-in-Medicine-and-Biology-Society (EMBC)

作者： Cayir, Sercan Yetik, Imam Samil TOBB Univ Econ & Technol Dept Elect & Elect Engn TR-06560 Ankara Turkey

ISBN: (纸本)9781509028092

Laser-assisted hair removal devices aim to remove body hair permanently. In most cases, these devices irradiate the whole area of the skin with a homogenous power density. Thus, a significant portion of the skin, where hair is not present, is burnt unnecessarily causing health risks. Therefore, methods that can distinguish hair regions automatically would be very helpful avoiding these unnecessary applications of laser. This study proposes a new system of algorithms to detect hair regions with the help of a digital camera. Unlike previous limited number of studies, our methods are very fast allowing for real-time application. Proposed methods are based on certain features derived from histograms of hair and skin regions. We compare our algorithm with competing methods in terms of localization performance and computation time and show that a much faster real-time accurate localization of hair regions is possible with the proposed method. Our results show that the algorithm we have developed is extremely fast (around 45 milliseconds) allowing for real-time application with high accuracy hair localization (96.48 %).

关键词： Laser-Assisted Hair Removal image processing automated hair localization

来源：评论

学校读者我要写书评

暂无评论

Compressive Light Field Reconstructions using Deep Learning 30

Compressive Light Field Reconstructions using Deep Learning

引用

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

作者： Gupta, Mayank Jauhari, Arjun Kulkarni, Kuldeep Jayasuriya, Suren Molnar, Alyosha Turaga, Pavan Arizona State Univ Tempe AZ 85287 USA Cornell Univ Ithaca NY 14853 USA Carnegie Mellon Univ Pittsburgh PA 15213 USA

ISBN: (纸本)9781538607336

Light field imaging is limited in its computational processing demands of high sampling for both spatial and angular dimensions. Single-shot light field cameras sacrifice spatial resolution to sample angular viewpoints, typically by multiplexing incoming rays onto a 2D sensor array. While this resolution can be recovered using compressive sensing, these iterative solutions are slow in processing a light field. We present a deep learning approach using a new, two branch network architecture, consisting jointly of an autoencoder and a 4D CNN, to recover a high resolution 4D light field from a single coded 2D image. This network decreases reconstruction time significantly while achieving average PSNR values of 26-32 dB on a variety of light fields. In particular, reconstruction time is decreased from 35 minutes to 6.7 minutes as compared to the dictionary method for equivalent visual quality. These reconstructions are performed at small sampling/compression ratios as low as 8%, allowing for cheaper coded light field cameras. We test our network reconstructions on synthetic light fields, simulated coded measurements of real light fields captured from a Lytro Illum camera, and real coded images from a custom CMOS diffractive light field camera. The combination of compressive light field capture with deep learning allows the potential for real-time light field video acquisition systems in the future.

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

12th International conference on Intelligent Information Hiding and Multimedia Signal processing, IIH-MSP 2016

12th International Conference on Intelligent Information Hid...

引用

12th International conference on Intelligent Information Hiding and Multimedia Signal processing, IIH-MSP 2016

ISBN: (纸本)9783319502083

The proceedings contain 40 papers. The special focus in this conference is on Information Hiding, Secret Sharing, Speech Signal processing, Communication Protocols, Techniques, Encryption and Authentication Methods. The topics include: A revisit to LSB substitution based data hiding for embedding more information;behavior steganography in social network;robust steganography using texture synthesis;a quantization-based image watermarking scheme using vector dot product;high-capacity robust watermarking approach for protecting ownership right;a data hiding method based on multi-predictor and pixel value ordering;a large payload webpage data embedding method using CSS attributes modification;the study of steganographic algorithms based on pixel value difference;digital audio watermarking robust against locality sensitive hashing;copyright protection method based on the main feature of digital images;a study on tailor-made speech synthesis based on deep neural networks;an improved 5-2 channel downmix algorithm for 3D audio reproduction;investigation on the head-related modulation transfer function for monaural DOA;temporal characteristics of perceived reality of multimodal contents;research on frequency automatically switching technology for china highway traffic radio;an automatic decoding method for Morse signal based on clustering algorithm;a novel digital rights management mechanism on peer-to-peer streaming system;a framework for supporting application level interoperability between IPv4 and IPv6;a new image encryption instant communication method based on matrix transformation and a three-party password authenticated key exchange protocol resistant to stolen smart card attacks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Additive Noise Removal by Combining Non Local Means Filtering and a Local Fuzzy Filter - A Fusion Approach 1st

Additive Noise Removal by Combining Non Local Means Filterin...

引用

1st International conference on Advances in Computing and Data Sciences (ICACDS)

作者： Raju, G. Wahid, Farha Fatina Sugandhi, K. Kannur Univ Dept Informat Technol Kannur Kerala India

ISBN: (纸本)9789811054273;9789811054266

Additive noise is one among the prominent types of noises which degrades the quality of images. A very large number of algorithms, in spatial, frequency and wavelet domain have been proposed to enhance images corrupted with additive noise. All the methods suggested have their own advantages as well as disadvantages. With the availability of parallel processing capability, in low end workstations and systems, fusion of two or more de-noising methods has become a topic of interest. In this paper, we have implemented one of the recent contributions to mean filter - a fuzzy filter. Also, as a complementary filter, the basic Non Local Means filter is implemented. Experiments were carried out by fusing the results obtained through the two filters. The results obtained establish the merit of the fusion approach.

关键词： Additive noise image de-noising image fusion NLM Fuzzy filter

来源：评论

学校读者我要写书评

暂无评论

LCDet: Low-Complexity Fully-Convolutional Neural Networks for Object Detection in Embedded systems 30

LCDet: Low-Complexity Fully-Convolutional Neural Networks fo...

引用

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

作者： Tripathi, Subarna Dane, Gokce Kang, Byeongkeun Bhaskaran, Vasudev Truong Nguyen Univ Calif San Diego La Jolla CA 92093 USA Qualcomm Inc San Diego CA USA

ISBN: (纸本)9781538607336

Deep Convolutional Neural Networks (CNN) are the state-of-the-art performers for the object detection task. It is well known that object detection requires more computation and memory than image classification. In this work, we propose LCDet, a fully-convolutional neural network for generic object detection that aims to work in embedded systems. We design and develop an end-to-end TensorFlow(TF)-based model. The detection works by a single forward pass through the network. Additionally, we employ 8-bit quantization on the learned weights. As a use case, we choose face detection and train the proposed model on images containing a varying number of faces of different sizes. We evaluate the face detection performance on publicly available dataset FDDB and Widerface. Our experimental results show that the proposed method achieves comparative accuracy comparing with state-of-the-art CNN-based face detection methods while reducing the model size by 3x and memory-BW by 3 - 4x comparing with one of the best real-time CNN-based object detector YOLO [23]. Our 8-bit fixed-point TF-model provides additional 4x memory reduction while keeping the accuracy nearly as good as the floating point model and achieves 20x performance gain compared to the floating point model. Thus the proposed model is amenable for embedded implementations and is generic to be extended to any number of categories 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

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：