In this work we propose a fully end-to-end approach for multi-spectral image registration and fusion. Our fusion method combines images from different spectral channels into a single fused image using approaches for l...
详细信息
In this work we propose a fully end-to-end approach for multi-spectral image registration and fusion. Our fusion method combines images from different spectral channels into a single fused image using approaches for low and high frequency signals. A prerequisite of fusion is the geometric alignment between the spectral bands, commonly referred to as registration. Unfortunately, common methods for image registration of a single spectral channel might prove inaccurate on images from different modalities. For that end, we introduce a new algorithm for multi-spectral image registration, based on a novel edge descriptor of feature points. Our method achieves an accurate alignment allowing us to further fuse the images. It is experimentally shown to produce a high quality of multi-spectral image registration and fusion under challenging scenarios.
Mobile robots are valuable educational tools due to they raise enthusiasm in students and also given the diversity of technological disciplines that they involve, including electronics, mechanics, sensing systems, emb...
详细信息
Mobile robots are valuable educational tools due to they raise enthusiasm in students and also given the diversity of technological disciplines that they involve, including electronics, mechanics, sensing systems, embedded systems, signal and imageprocessing, wireless communication, programming and computational algorithms, and artificial intelligence, among others. The present work describes a prototype of an open hardware wheeled mobile robot for educational purposes -the EduRoMAA. This robot is designed to be used as an experimental platform in embedded system programming courses, covering from the initial to advanced level, using different processing boards such as: Arduino and EduCIAA. A detailed description of the mechanical and electronic parts, as well as a PC application for testing the robot is presented. Tested processing boards are described, including their pros and cons.
In bridge health monitoring, the detection and localization of surface defects are highly important for health condition evaluation. Due to the limitation of manual detection, it is easier to measure those defects in ...
详细信息
ISBN:
(纸本)9781538670774;9781538670767
In bridge health monitoring, the detection and localization of surface defects are highly important for health condition evaluation. Due to the limitation of manual detection, it is easier to measure those defects in a more automatic way. Machine learning is a hot topic in the recent decade, and the contribution of Artificial Neural Network (ANN) is especially remarkable, which is the most widely used models of machine learning in the image-processing field. In this paper, we will discuss two ANN-based algorithms (Back propagation (BP) and Self-Organizing Maps (SOM)) and their applications for the recognition of surface defect on images taken from bridges. Moreover, a combined network algorithm with BP and SOM is designed in order to improve the performance in crack image segmentation, and analysis over this network is carried out specifically.
This paper is devoted to present technique of the use of imageprocessing for lab-on-a-chip techniques. algorithms and methods for cell detecting, obtaining their parameters and multiparametric cell tracking in lab-on...
详细信息
ISBN:
(纸本)9788363578121
This paper is devoted to present technique of the use of imageprocessing for lab-on-a-chip techniques. algorithms and methods for cell detecting, obtaining their parameters and multiparametric cell tracking in lab-on-a-chip were presented and discussed from the point of real-time detection.
A recent and fascinating interest in computational symmetry for computer vision and computer graphics applications has led to a remarkable realization of new symmetry detection algorithms. Such a concern is culminated...
详细信息
ISBN:
(纸本)9781509063444
A recent and fascinating interest in computational symmetry for computer vision and computer graphics applications has led to a remarkable realization of new symmetry detection algorithms. Such a concern is culminated in a symmetry detection competition as a workshop affiliated with the 2011 and 2013 CVPR conferences. In this paper, we propose a method based on the computation of the symmetry level associated to each pixel. Such a value is determined through the energy balance of the even/odd decomposition of a patch with respect to a central axis (which is equivalent to estimate the middle point of a row-wise convolution). Peaks localization along the perpendicular direction of each angle allows to identify possible symmetry axes. The evaluation of a feature based on gradient information allows to establish a classification confidence for each detected axis. By adopting the aforementioned rigorous validation framework, the proposed method indicates significant performance increase.
The proceedings contain 38 papers. The topics discussed include: a system for exploratory analysis in cloud;privacy-preserving and unforgeable searchable encrypted audit logs for cloud storage;a real-time sensor netwo...
ISBN:
(纸本)9781538658505
The proceedings contain 38 papers. The topics discussed include: a system for exploratory analysis in cloud;privacy-preserving and unforgeable searchable encrypted audit logs for cloud storage;a real-time sensor network aggregation computing system;a distributed technique for repairing inconsistencies of functional dependencies;MTComm based virtualization and integration of physical machine operations with digital-twins in cyber-physical manufacturing cloud;difference function projective synchronization for secure communication based on complex chaotic systems;an approach of fog detecting magnitude using referenceless perceptual image defogging;reliability ranking prediction for cloud services via skyline;a cooperative game for online cloud federation formation based on security risk assessment;a knowledge-based approach to reducing complexity of maintaining semantic constraints in data exchange and integration;review on big data fusion methods of quality inspection for consumer goods;unique topic query processing on cloud;enterprise workflow modeling based on priced timed Petri nets;performance evaluation of BATMAN-adv wireless mesh network routing algorithms;FASTBEE: a fast and self-adaptive clustering algorithm towards to edge computing;cryptanalysis of an RFID ownership transfer protocol based on cloud;network security based on d-s evidence theory optimizing CS-BP neural network situation assessment;access control scheme based on combination of blockchain and XOR-coding for ICN;and an effective scheme to detect and prevent tampering on the physical layer of WSN.
The proceedings contain 31 papers. The special focus in this conference is on Web and Big Data. The topics include: Text Classification Methods Based on SVD and FCM;An imageprocessing Method via OpenCL for Identifica...
ISBN:
(纸本)9783030012977
The proceedings contain 31 papers. The special focus in this conference is on Web and Big Data. The topics include: Text Classification Methods Based on SVD and FCM;An imageprocessing Method via OpenCL for Identification of Pulmonary Nodules;Pulmonary Nodule Segmentation Method of CT images Based on 3D-FCN;research on video Recommendation Algorithm Based on Knowledge Reasoning of Knowledge Graph;a Hybrid Framework for Query processing and Data Analytics on Spark;improving Network-Based Top-N Recommendation with Background Knowledge from Linked Open Data;a Web-Based Theme-Related Word Set Construction Algorithm;classifying Personal Photo Collections: An Event-Based Approach;a Learning Analytic Model for Smart Classroom;a Self-representation Model for Robust Clustering of Categorical Sequences;understanding User Interests Acquisition in Personalized Online Course Recommendation;a Learning Analytics System for Cognition Analysis in Online Learning Community;A Semantic Role Mining and Learning Performance Prediction Method in MOOCs;MOOC Guider: An End-to-End Dialogue System for MOOC Users;AUnet: An Unsupervised Method for Answer Reliability Evaluation in Community QA systems;model and Practice of Crowd-Based Education;Exploring Business Models and Dynamic Pricing Frameworks for SPOC Services;speed-Up algorithms for Happiness-Maximizing Representative Databases;multi-location Influence Maximization in Location-Based Social Networks;emotion Analysis for the Upcoming Response in Open-Domain Human-Computer Conversation;diversified Spatial Keyword Query on Topic Coverage;a Recurrent Neural Network Language Model Based on Word Embedding;inferring Social Ties from Multi-view Spatiotemporal Co-occurrence;sequence-As-Feature Representation for Subspace Classification of Multivariate Time Series;discovering Congestion Propagation Patterns by Co-location Pattern Mining;Spectroscopy-Based Food Internal Quality Evaluation with XGBoost Algorithm.
The present paper focuses on a multimodal system based on electrooculography, imageprocessing, and speech recognition for simple and real-time controlling of robotic systems such as manipulators. Electrooculogram was...
详细信息
ISBN:
(纸本)9781538626405
The present paper focuses on a multimodal system based on electrooculography, imageprocessing, and speech recognition for simple and real-time controlling of robotic systems such as manipulators. Electrooculogram was based on voluntary eye movements and processed by thresholds;the average accuracy obtained was about 75 %. Two real-time and simple algorithms are presented for imageprocessing and speech recognition. Color segmentation successfully recognizes red, green, and blue objects. The algorithm used for speech processing represents the correlation of frequency domain for the desired word with the source words. The average accuracy of this algorithm is 90%. It is used for achieving higher degrees of freedom and commanding to the system in order for confirmation and next step recognition. Four different multimodal architectures were designed as combinations of electrooculography, imageprocessing, and speech processingalgorithms. Simulation of the architectures combines human vision (EOG) and robot vision (imageprocessing) in one system to improve accuracy, operation speed, degrees of freedom, orientation efficiency, target recognition, and the accuracy of robot's gripper distance adjustment. At last one of the architectures is implemented on a two degree of-freedom robotic manipulator to evaluate system operation speed and capability. The results indicated that the architectures control the system in a highly efficient manner. It can be thus concluded that a multimodal system based on EOG and imageprocessing improves the accuracy and reliability of rehabilitation and industrial systems with high operation speed.
A key challenge when displaying and processing sensed real-time 3D data is efficiency of generating and post-processingalgorithms in order to acquire high quality 3D content. In contrast, our approach focuses on volu...
详细信息
A key challenge when displaying and processing sensed real-time 3D data is efficiency of generating and post-processingalgorithms in order to acquire high quality 3D content. In contrast, our approach focuses on volumetric generation and processing volumetric data using an efficient low-cost hardware setting. Acquisition of volumetric data is performed by connecting several Kinect v2 scanners to a single PC that are subsequently calibrated using planar pattern. This process is by no means trivial and requires well designed algorithms for fast processing and quick rendering of volumetric data. This can be achieved by fusing efficient filtering methods such as Weighted median filter (WM), Radius outlier removal (ROR) and Laplace-based smoothing algorithm. In this context, we demonstrate the robustness and efficiency of our technique by sensing several scenes.
Denoising using Discrete Wavelet Transform is very popular in imageprocessing. The main tasks of this kind of approach include the image decomposition into random sub-bands and the Wavelet coefficients extraction. In...
详细信息
ISBN:
(纸本)9781538693865;9781538693858
Denoising using Discrete Wavelet Transform is very popular in imageprocessing. The main tasks of this kind of approach include the image decomposition into random sub-bands and the Wavelet coefficients extraction. In this paper a GPU parallel algorithm to compute and extract Wavelet coefficients is proposed. As preprocessing we use the Donoho method to study the amount of noise present in the images; while for suppression or shrinkage criterion we suggest the Garrote threshold. The selection process in this work was performed through two algorithms: Hard and Non-Negative Garrote Thresholding. Our work-frame uses a parallel approach to extract Wavelet coefficients at level of sub-bands, so that significant improvements are achieved over the CPU version, which a speedup gain of around four times with respect to the serial version. Some tests and numerical experiments show the efficiency and the quality obtained using the proposed method.
暂无评论