检索结果-内蒙古大学图书馆

International Conference on Signal Processing Proceedings (ICSP)

作者： Xin Li Pengyu Guo Hongliang Zhang Xiaohu Zhang Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation National University of Defense Technology Changsha China

ISBN: (纸本)9781479921898

In order to determine the 3D motion of a point target from a monocular moving camera, a locating method is proposed by taking advantage of the property that the positions of the target in a short time are relative to each other and the kernel trick of least squares support vector machine (LS-SVM). This method has better noise immunity due to its robust estimation property, and has high accuracy as the kernel function added to the trajectory reconstruction, and has high efficiency because of the incremental and decremental technique. To illustrate the performance of the method employed in the paper, contrastive simulation experiment on location error and execution time is designed. The results show the accuracy and efficiency of the method. This method is applicable for high precision measurement of motion targets by passive observations on various mobile platforms.

关键词： Trajectory Cameras Support vector machines Computational modeling Mathematical model Polynomials Motion measurement

来源：评论

学校读者我要写书评

暂无评论

Airborne vision-aided landing navigation system for fixed-wing UAV

Airborne vision-aided landing navigation system for fixed-wi...

引用

International Conference on Signal Processing Proceedings (ICSP)

作者： Pengyu Guo Xin Li Yang Gui Xiang Zhou Hongliang Zhang Xiaohu Zhang Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation National University of Defense Technology Changsha China

ISBN: (纸本)9781479921898

In order to achieve accurate automatic recovery of fixed-wing UAV at the known landing site, an airborne aided landing navigation system based on machine vision is designed and realized, and the key algorithms of the system are investigated. First, according to the need of automatic landing, such as high precision, real-time and strong anti-interference ability, a scheme of vision-aided landing navigation system is proposed based on guidance illuminant with fixed waveband. Then, aiming at the characteristics of the system, the primary algorithms are designed and realized including camera system calibration with optical filter, fast and accurate blob location, and pose estimation for quasisingular case. Finally, the system is established in DSP, and verified by the experiment. Feasibility experimental results indicate the tendency of the location data between GPS and the system is consistent. Precision experimental results show the extraction precision of blob centroid in the image coordinates is better than 0.1 subpixel, and the localization precisions of UAV along the runway, vertical to the runway and the height orientation are 0.234m, 0.025m and 0.009m when UAV is at about 100m away from the ideal landing point. Efficiency experimental results illustrate the processing frequency is 100Hz~15Hz. It can satisfy the requirements of automatic landing which are real-time, high precision as well as strong anti-interference ability.

关键词： navigation Cameras Calibration Accuracy Optical filters Machine vision Digital signal processing

来源：评论

学校读者我要写书评

暂无评论

Rigid Body Pose Estimation from Line Correspondences

Rigid Body Pose Estimation from Line Correspondences

引用

International Conference on image, vision and Computing (ICIVC)

作者： Yantao Yue Xiangyi Sun College of Aerospace Science and Engineering Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation Changsha China

ISBN: (纸本)9781538649923

In this paper, we aim to solve pose estimation of rigid body motion in real time with 3d lines model. According to the line's perspective projection model, we design a new error function expressed by the average integral of the distance between line segments to estimate parameters. Considering the continuely of motion, we restore cracked line segements with re-projection of Model lines Constrianted. Last, we proposal to estimate many frames jointly in framework of SFM and get better precious while bears slow speed. Comparisons on synthetic and real images demonstrate that baseline methods get accuracy estimations in complex environments. For plane objects, the precious of pose on x, y, z axes are better than 0.5m in 100m distance, and those of relative positions perpendicular to the optical axis and along the optical axis are better than 0.3%.

关键词： Three-dimensional displays image segmentation Solid modeling Optimization Estimation Mathematical model Cameras

来源：评论

学校读者我要写书评

暂无评论

Deformable Patch-based NCC Measure for Visual Tracking

Deformable Patch-based NCC Measure for Visual Tracking

引用

2016 International Conference on Computer, Mechatronics and Electronic Engineering (CMEE 2016)

作者： Xian-guo YU Qi-feng YU Liang XIE Jin-bo LIU College of Aerospace Science and Engineering National University of Defense Technology Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation National University of Defense Technology

This paper tries to solve the visual tracking problem in a simple and intuitive way. In particular, we extend the traditional NCC measure with deformable part-based models(DPM) to get a new image similarity measurement and feed it into the particle filtering framework for visual object tracking. The proposed target model consists of a root template and a set of patch templates. The root is the whole target image, and patches are sub-regions of the target image. NCC responses are calculated for the root template with test images, and every patch template with their corresponding sub-regions of test images. The likelihood between the target and a test image is calculated by integrating all template responses. Instead of directly accumulating all responses, we employ the idea of DPM that combines global and local responses using an elastic model to get a robust image likelihood measure. We also explore an exhaustive patch representation mechanism by introducing the dense NCC feature map. The proposed feature map enables the usage of all patches in the target region for target description and similarity measure. Finally, we provide both qualitative and quantitative experiments against state-of-the-arts on the TB50 dataset. Experimental results show that the proposed similarity measure is effective for visual tracking under challenging scenarios.

关键词： Visual tracking Deformable patches Similarity measurement

来源：评论

学校读者我要写书评

暂无评论

Oriented Object Detection in Optical Remote Sensing images using Deep Learning: A Survey

arXiv

引用

arXiv 2023年

作者： Wang, Kun Wang, Zi Li, Zhang Su, Ang Teng, Xichao Liu, Minhao Yu, Qifeng College of Aerospace Science and Engineering National University of Defense Technology Changsha410000 China Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation Changsha410000 China

Oriented object detection is one of the most fundamental and challenging tasks in remote sensing, aiming to locate and classify objects with arbitrary orientations. Recent years have witnessed remarkable progress in oriented object detection using deep learning techniques. Given the rapid development of this field, this paper aims to provide a comprehensive survey of recent advances in oriented object detection. To be specific, we first review the technical evolution from horizontal object detection to oriented object detection and summarize the specific challenges, including feature misalignment, spatial misalignment, and periodicity of angle. Subsequently, we further categorize existing methods into detection framework, oriented bounding box (OBB) regression, and feature representations, and discuss how these methods address the above challenges in detail. In addition, we cover several publicly available datasets and performance evaluation protocols. Furthermore, we provide a comprehensive comparison and analysis of state-of-the-art oriented object detection methods. Toward the end of this paper, we discuss several future directions for oriented object detection. © 2023, CC BY-NC-ND.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Solar Wings' 3D Deformation measurement of Videometrics System

Solar Wings' 3D Deformation Measurement of Videometrics Syst...

引用

2016 International Conference on Mechatronics Engineering and Information Technology(ICMEIT 2016)

作者： Fengzhe Zhang Jiexin Zhou Guangcheng Dai Yijie Bian Xia Yang College of Aerospace science and Engineering National University of Defense Technology Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation College of Computer National University of Defense Technology

ISBN: (纸本)9781510829039

In this paper, we proposed a novel videometrics system application based on binocular stereovision. And we successfully measured three-dimensional(3D) deformation of the solar wings plane. Videometrics has its inherent advantages in solving these problems, which can provide a non-contact, automatic, dynamic and high precision measurement. This system mainly includes camera calibration module, image feature point matching module and 3D model display module. Camera calibration is a key technology of stereoscopic calibration. We applied Zhang Zhengyou new approach, which is widely used to calibrate the camera. For the image matching part, we used the correlation matching method and Ackerman Least Squares method, which can achieve accuracy of 1/10 or even 1/100 pixel. To further optimize the process, epipolar constraint was adopted. Two-dimensional search can be converted into a one-dimensional search, thereby highly reducing the matching computation. Moreover, 3D space measured results obtained in the target point test can be displayed in the system. With the target region segmentation statistics, the correct result is more vividly displayed and accurately measured. Based on the proposed method we have designed the measurement software and simulate the solar wings' s 3D deformation measurement experiment. The re-projection error for both of stereo cameras is no more than 0.2 pixel. The height difference of solar wings plane is 0.014 mm. The method proposed in this paper has significance both on the theory and application.

关键词： Solar wings Videometrics Binocular stereovision Camera Calibration Least squares scene matching Epipolar constraint

来源：评论

学校读者我要写书评

暂无评论

Accurate Monocular Pose measurement for Aircraft Transportation Monitoring

Accurate Monocular Pose Measurement for Aircraft Transportat...

引用

IEEE International Conference on Unmanned Systems (ICUS)

作者： Xiaoliang Sun Zhuo Zhang Daoming Bi Liangchao Guo Qiufu Wang Lin Chen Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation College of Aerospace Science and Engineering National University of Defense Technology Changsha China AVIC Shenyang Aircraft Design & Research Institute Shenyang China

ISBN: (数字)9798350384185

ISBN: (纸本)9798350384192

Comprehensive statue monitoring is of great significance for aircraft safety and efficient transportation. This paper proposes an accurate and efficient monocular pose measurement method for aircraft transportation monitoring. The proposed method firstly models the aircraft using a sparse keypoint set for simplicity. Then, based on the simple representation, a trained neural network model is adapted to detect the keypoints from the input monocular image to build 2D-3D correspondences for pose measurement. Coarse pose estimation is got by solving a Perspective-n-Point problem based on the 2D-3D correspondences. For accurate pose measurement, this paper uses the contour feature to refine the pose measurement. The refinement is achieved via gradient orientation based template matching. Experimental results indicate that the proposed method achieves accurate and efficient monocular pose measurement for aircraft transportation monitoring.

关键词： YOLO Adaptation models Accuracy Three-dimensional displays Atmospheric modeling Pose estimation Transportation Aircraft manufacture Aircraft Monitoring

来源：评论

学校读者我要写书评

暂无评论

Automatic detection of breast lesions in automated 3D breast ultrasound with cross-organ transfer learning

引用

虚拟现实与智能硬件（中英文） 2024年第3期6卷 239-251页

作者： Lingyun BAO Zhengrui HUANG Zehui LIN Yue SUN Hui CHEN You LI Zhang LI Xiaochen YUAN Lin XU Tao TAN Affiliated Hangzhou First People's Hospital School of MedicineWestlake UniversityHangzhou 330100China Faculty of Applied Sciences Macao Polytechnic UniversityMacao 999078China Pathology Department Changsha First HospitalChangsha 410073China Radiology Department Changsha First HospitalChangsha 410073China College of Aerospace Science and Engineering National University of Defense TechnologyChangsha 410073China Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation Changsha 410073China School of Information Science and Technology Shanghaitech UniversityShanghai 201210China

Background Deep convolutional neural networks have garnered considerable attention in numerous machine learning applications,particularly in visual recognition tasks such as image and video *** is a growing interest in applying this technology to diverse applications in medical image *** three dimensional Breast Ultrasound is a vital tool for detecting breast cancer,and computer-assisted diagnosis software,developed based on deep learning,can effectively assist radiologists in ***,the network model is prone to overfitting during training,owing to challenges such as insufficient training *** study attempts to solve the problem caused by small datasets and improve model detection *** We propose a breast cancer detection framework based on deep learning(a transfer learning method based on cross-organ cancer detection)and a contrastive learning method based on breast imaging reporting and data systems(BI-RADS).Results When using cross organ transfer learning and BIRADS based contrastive learning,the average sensitivity of the model increased by a maximum of 16.05%.Conclusion Our experiments have demonstrated that the parameters and experiences of cross-organ cancer detection can be mutually referenced,and contrastive learning method based on BI-RADS can improve the detection performance of the model.

关键词： Breast ultrasound Automated 3D breast ultrasound Breast cancers Deep learning Transfer learning Convolutional neural networks Computer-aided diagnosis Cross organ learning

来源：评论

学校读者我要写书评

暂无评论

Bridging the Domain Gap in Satellite Pose Estimation: a Self-Training Approach based on Geometrical Constraints

arXiv

引用

arXiv 2022年

作者： Wang, Zi Chen, Minglin Guo, Yulan Li, Zhang Yu, Qifeng The College of Aerospace Science and Engineering National University of Defense Technology Changsha410073 China Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation Changsha410073 China The School of Electronics and Communication Engineering Sun Yat-Sen University Shenzhen518000 China The College of Electronic Science and Technology National University of Defense Technology Changsha410073 China

Recently, unsupervised domain adaptation in satellite pose estimation has gained increasing attention, aiming at alleviating the annotation cost for training deep models. To this end, we propose a self-training framework based on the domain-agnostic geometrical constraints. Specifically, we train a neural network to predict the 2D keypoints of a satellite and then use PnP to estimate the pose. The poses of target samples are regarded as latent variables to formulate the task as a minimization problem. Furthermore, we leverage fine-grained segmentation to tackle the information loss issue caused by abstracting the satellite as sparse keypoints. Finally, we iteratively solve the minimization problem in two steps: pseudo-label generation and network training. Experimental results show that our method adapts well to the target domain. Moreover, our method won the 1st place on the sunlamp task of the second international Satellite Pose Estimation Competition. Copyright © 2022, The Authors. All rights reserved.

关键词： Satellites

来源：评论

学校读者我要写书评

暂无评论

An Optimal Design Scheme of Missile Trajectory

引用

Journal of Physics: Conference Series 2022年第1期2220卷

作者： Shilin Zhang College of Aerospace Science and Engineering National University of Defense Technology Changsha Hunan 410073 Hunan Provincial Key Laboratory of Image Measurement and Vision Navigation Changsha Hunan 410073

The optimization of missile trajectory is an important part of missile design. Its difficulty lies in the complex force of the missile and the different laws of the trajectory parameters in each stage. The existing trajectory simulation models have problems such as complex models, low accuracy, simplifying the initial motion state of the missile, and failing to fully reflect the characteristics of the active and passive motions. In order to solve these problems, this paper extracts important model variables on the basis of analyzing the force of the missile at different stages, and establishes a ballistic model with time-varying parameters. At the same time, according to the engineering design method of the missile flight program, the change rule of the pitch angle of the active phase of the missile with time is drawn up. Finally, an intercontinental ballistic missile is used as the object to simulate and analyze the proposed model. The results show that the method proposed in this paper can design and optimize missile trajectory well.

关键词：

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：