检索结果-内蒙古大学图书馆

An Unsupervised moving object detection Network for UAV Videos

DRONES 2025年第2期9卷 150-150页

作者： Fan, Xuxiang Wen, Gongjian Gao, Zhinan Chen, Junlong Jian, Haojun Natl Univ Def Technol Coll Elect Sci & Technol Changsha 410073 Peoples R China

UAV moving object detection focuses on identifying moving objects in images captured by UAVs, with broad applications in regional surveillance and event reconnaissance. Compared to general moving object detection scenarios, UAV videos exhibit unique characteristics, including foreground sparsity and varying target scales. The direct application of conventional background modeling or motion segmentation methods from general settings may yield suboptimal performance in UAV contexts. This paper introduces an unsupervised UAV moving object detection network. Domain-specific knowledge, including spatiotemporal consistency and foreground sparsity, is integrated into the loss function to mitigate false positives caused by motion parallax and platform movement. Multi-scale features are fully utilized to address the variability in target sizes. Furthermore, we have collected a UAV moving object detection dataset from various typical scenarios, providing a benchmark for this task. Extensive experiments conducted on both our dataset and existing benchmarks demonstrate the superiority of the proposed algorithm.

关键词： moving object detection UAV optical flow motion segmentation

来源：评论

学校读者我要写书评

暂无评论

Real-Time moving object detection for 3-D LiDAR Using Occlusion Accumulation in Range Image

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2025年 74卷

作者： Kim, Junha Kim, Haram Oh, Changsuk Kim, Changhyeon Kim, H. Jin Seoul Natl Univ Dept Mech & Aerosp Engn Seoul 08826 South Korea Seoul Natl Univ Automat & Syst Res Inst ASRI Seoul 08826 South Korea 42dot Seoul 13449 South Korea Seoul Natl Univ Artificial Intelligence Inst AIIS Seoul 08826 South Korea Samsung Res Seoul 06765 South Korea Seoul Natl Univ Dept Aerosp Engn Seoul 08826 South Korea

Deep learning methods have been applied to detect moving objects from 3-D LiDAR data, but they require extra computing devices, such as graphics processing units (GPUs), and often struggle to operate at real-time LiDAR frame rates. In addition, retraining is necessary for new environments, which demands significant resources. Nonlearning methods that operate online without prior maps and rely solely on a central processing unit (CPU) often fail to achieve real-time performance. Furthermore, they are limited in applicability due to their dependence on specific pose estimation algorithms. To address these issues, we propose a novel occlusion accumulation framework in the range image domain for real-time 3-D LiDAR moving object detection on a CPU. By incorporating compensation strategies that account for LiDAR artifacts and measurement sparsity, the proposed method reduces false positives (FPs) and improves detection accuracy. Our approach is also flexible, integrating seamlessly with different pose estimation algorithms without performance degradation. Extensive experiments on KITTI, Apollo, and CARLA datasets demonstrate that the proposed method achieves competitive results compared to state-of-the-art (SOTA) learning-based methods on KITTI and outperforms them in cross-validation on Apollo and CARLA. Moreover, the proposed method operates four times faster than SOTA nonlearning-based methods while offering higher detection accuracy, making it highly suitable for real-world applications in autonomous driving and robotic navigation. We provide all the source code and datasets used in this article to the public at: https://***/JunhaAgu/Mapless_moving.

关键词： 3-D LiDAR moving object detection occlusion accumulation occlusion accumulation range image range image range image

来源：评论

学校读者我要写书评

暂无评论

An efficient representative coefficient total variation method for moving object detection in noisy videos

引用

SIGNAL IMAGE AND VIDEO PROCESSING 2025年第7期19卷 1-14页

作者： Zhu, Hui Feng, Xiangchu Xidian Univ Sch Math & Stat Xian 710071 Peoples R China

Traditional low-rank sparse decomposition (LRSD) methods have difficulty in detecting moving objects quickly and accurately in noisy videos due to the high-dimensional singular value decomposition (SVD) of the video data and noise. Noise contaminates both the video background and the moving object. Therefore, there is an urgent need to remove the noise from both the background and the moving object, as well as to improve the efficiency of the traditional methods for detecting moving objects. To tackle these challenges, we propose a novel low-rank sparse decomposition method for processing noisy videos. Our approach integrates representative coefficient total variation (RCTV) regularization and spatio-temporal total variation (STTV) regularization to reconstruct the video background and detect moving objects. Specifically, RCTV regularization is applied to constrain the low-rank video background, effectively eliminating background noise and reducing the computational complexity of traditional model optimization. Additionally, STTV regularization enforces the smooth sparsity of moving objects, significantly mitigating false detections caused by dynamic background variations and noise. Experimental results demonstrate that the proposed method successfully recovers a clean background while accurately detecting moving objects in noisy videos.

关键词： moving object detection Spatio-temporal continuity Video denoising Representative coefficient total variation

来源：评论

学校读者我要写书评

暂无评论

moving object detection using unstable camera for video surveillance systems

引用

OPTIK 2015年第20期126卷 2436-2441页

作者： Lee, Seungwon Kim, Nahyun Jeong, Kyungwon Park, Kyungju Paik, Joonki Chung Ang Univ Dept Image Seoul 156756 South Korea

This paper presents a robust moving object detection method by compensating for motion of an unstable camera. Assuming that global camera motion results in affine transform between two successive frames, local affine motions are separately estimated in multiple pre-specified regions for fast, robust estimation of the global motion. The global camera motion is then estimated by the least squares method using the pre-estimated multiple local affine motions. Given a current frame as the reference, the subsequent frame is registered to the current frame using the estimated global motion. The moving objects are finally detected using difference of Gaussian and non-parametric kernel density estimation from the set of registered three frames. Experimental results show that the proposed method can robustly detect moving objects in unstable imaging environment for intelligent surveillance systems using various types of cameras including pan-tilt-zoom (PTZ) and unmanned aerial vehicle (UAV) cameras. (C) 2015 Elsevier GmbH. All rights reserved.

关键词： moving object detection Surveillance system Camera motion estimation Unstable camera

来源：评论

学校读者我要写书评

暂无评论

moving object detection Method via ResNet-18 With Encoder-Decoder Structure in Complex Scenes

引用

IEEE ACCESS 2019年 7卷 108152-108160页

作者： Ou, Xianfeng Yan, Pengcheng Zhang, Yiming Tu, Bing Zhang, Guoyun Wu, Jianhui Li, Wujing Hunan Inst Sci & Technol Sch Informat Sci & Engn Yueyang 414006 Peoples R China Hunan Inst Sci & Technol Machine Vis & Artificial Intelligence Res Ctr Yueyang 414006 Peoples R China

In complex scenes, dynamic background, illumination variation, and shadow are important factors, which make conventional moving object detection algorithms suffer from poor performance. To solve this problem, a moving object detection method via ResNet-18 with encoder-decoder structure is proposed to segment moving objects from complex scenes. ResNet-18 with encoder-decoder structure possesses pixel-level classification capability to divide pixels into foreground and background, and it performs well in feature extraction because of its layers are so shallow that many more low-scale features will be retained. First, the object frames and their corresponding artificial labels are input to the network. Then, feature vectors will be generated by the encoder, and they are converted into segmentation maps by the decoder through deconvolution processing. Third, a rough matching of the moving object regions will be obtained, and finally, the Euclidean distance is used to match the moving object regions accurately. The proposed method is suitable for the scenes where dynamic background, illumination variation, and shadow exist, and experimental results on the public standard CDnet2014 and I2R datasets, from both qualitative and quantitative comparison aspects, demonstrate that the proposed method outperforms state-of-the-art algorithms significantly, and its mean F-measure increased by 1.99%similar to 29.17%.

关键词： Complex scenes moving object detection ResNet-18 encoder-decoder network background subtraction

来源：评论

学校读者我要写书评

暂无评论

moving object detection via RPCA framework using non-convex low-rank approximation and total variational regularization

引用

SIGNAL IMAGE AND VIDEO PROCESSING 2023年第1期17卷 109-117页

作者： Chen, Tianfei Zhao, Dongliang Sun, Lijun Li, Shi Feng, Binbin Henan Univ Technol Minist Educ Key Lab Grain Informat Proc & Control Zhengzhou 450001 Peoples R China Henan Univ Technol Henan Key Lab Grain Photoelectr Detect & Control Zhengzhou 450001 Peoples R China Henan Univ Technol Coll Informat Sci & Engn Zhengzhou 450001 Peoples R China

moving object detection is one of the significant tasks in computer vision. Robust Principal Component Analysis (RPCA) is a common method for moving object detection. However, this algorithm fails to effectively utilize the low-rank prior information of the background and the spatio-temporal continuity of the foreground, and exhibits degraded performance in the existence of shadows, photometric variations, rapidly moving objects, dynamic backgrounds, and jitters. Therefore, a new model via RPCA framework using non-convex low-rank approximation and total variational regularization is proposed to detect moving objects. Firstly, this paper introduces a non-convex function to deal with the problem that the nuclear norm excessively penalizes large singular values for the sake of constraining the low-rank characteristics of video background availably. Then, the sparsity is strengthened with the l(1)-norm and the spatio-temporal continuity of the foreground is explored by TV regularization. Finally, the augmented Lagrange multiplier algorithm is expanded by the alternating direction multiplier strategy to solve the model. Extensive experiments show that the proposed method outperforms existing methods in terms of the accuracy of moving object detection and foreground extraction effect.

关键词： moving object detection Non-convex rank approximation Total variational regularization Robust principal component analysis

来源：评论

学校读者我要写书评

暂无评论

moving object detection based on incremental learning low rank representation and spatial constraint

引用

NEUROCOMPUTING 2015年 168卷 382-400页

作者： Dou, Jianfang Li, Jianxun Qin, Qin Tu, Zimei Shanghai Second Polytech Univ Sch Intelligent Mfg & Control Engn Dept Automat & Mech & Elect Engn Shanghai 201209 Peoples R China Shanghai Jiao Tong Univ Dept Automat Shanghai 200240 Peoples R China Minist Educ China Key Lab Syst Control & Informat Proc Shanghai 200240 Peoples R China

Background modeling and subtraction, the task to detect moving objects in a scene, is an important step in video analysis. In this paper, we present a novel moving object detection method based on Online Low Rank Matrix Recovery and graph cut from monocular video sequences. First, use the K-SVD method to initialize the dictionary to construct the background model, perform foreground detection with augmented Lagrange multipliers (ALM) and refine foreground values by spatial smooth constraint to extract the background and foreground information;Then obtain the clusters of foreground and background respectively using mean shift clustering on the background and foreground information;Third, initialize the S/T Network with corresponding image pixels as nodes (except S/T node);Calculate the data and smoothness term of graph;Finally, use max flow/minimum cut to segmentation S/T network to extract the motion objects. Online dictionary learning is adopted to update the background model. Experimental results on indoor and outdoor videos demonstrate the efficiency of our proposed method. (C) 2015 Elsevier B.V. All rights reserved.

关键词： Background subtraction Low rank representation moving object detection Dictionary learning Mean shift Graph cut

来源：评论

学校读者我要写书评

暂无评论

moving object detection under different weather conditions using full-spectrum light sources

引用

PATTERN RECOGNITION LETTERS 2020年 129卷 205-212页

作者： Boukhriss, Rania Rebai Fendri, Emna Hammami, Mohamed Sfax Univ MIRACL ISIMS Sfax Tunisia Sfax Univ MIRACL FS Rd Sokra Sfax Tunisia

The moving object detection always remains an active field of research given the variety of challenges related to this topic. In fact, most of the challenges related to the low illumination and weather conditions (fog, snow, rain, etc.) remain unresolved and require more developments. In this paper, our intrinsic objective is to overcome these challenges using an effective moving object detection method. Unlike most works in the literature that use one of the two infrared or visible spectra independently, we proposed a moving object detection method based on background modeling using the Full-Spectrum Light Sources (FSLS-MOD). To better ensure the adaptability and independence of the moving object speeds and sizes, the principle of the inter-frame differences' methods is introduced in the background modeling stage. Furthermore, we applied a new strategy to switch between the spectra allowing us to benefit from the advantages of each spectrum and carry out a better moving object detection even in bad weather conditions. An experimental study by quantitative and qualitative evaluations proved the robustness and effectiveness of our proposed method of moving object detection using the switching strategy between full-spectrum light sources under different illuminations and weather conditions. (C) 2019 Elsevier B.V. All rights reserved.

关键词： Full-spectrum Light Sources Weather Conditions Classification moving object detection Thermal Infrared Camera

来源：评论

学校读者我要写书评

暂无评论

moving object detection by low rank approximation and l₁-TV regularization on RPCA framework

引用

JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION 2018年 56卷 188-200页

作者： Shijila, B. Tom, Anju Jose George, Sudhish N. Natl Inst Technol Calicut Dept Elect & Commun Engn Calicut 673601 Kerala India

The detection of moving objects and the subtraction of the scene background are significant tasks for intelligent video surveillance systems as it is one among the fundamental steps. Inspired by the challenging cases yet to be resolved in moving object detection (MOD), a new formulation is done to detect moving objects from video sequences based on Robust Principal Component Analysis (RPCA) principle by adopting the regularization of Total Variation (TV) norm using a convergent convex optimization algorithm. While the nuclear norm exploits the low-rank property of background, the sparsity is enhanced by the l(1)-norm and the foreground spatial smoothness is explored by TV regularization. The goodness of this method lies in the reduced computational complexity, quickness and on the superiority acquired in quantitative evaluation based on F-measure, Recall and Precision with respect to the state of the art methods. (C) 2018 Elsevier Inc. All rights reserved.

关键词： moving object detection Low rank recovery Background subtraction Robust principle component analysis

来源：评论

学校读者我要写书评

暂无评论

moving object detection in video sequence images based on an improved visual background extraction algorithm

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2020年第39-40期79卷 29663-29684页

作者： Zuo, Junhui Jia, Zhenhong Yang, Jie Kasabov, Nikola Xinjiang Univ Coll Informat Sci & Engn 666 Shengli Rd Tianshan Dist Urumqi Peoples R China Shanghai Jiao Tong Univ Inst Image Proc & Pattern Recognit 2-427 SEIEE Bldg800 Dong Chun Rd Shanghai Peoples R China Auckland Univ Technol Knowledge Engn & Discovery Res Inst Private Bag 92006 Auckland New Zealand

Visual background extraction is a method for detecting moving objects in video sequence images. In the traditional visual background extraction algorithm, ghost phenomena and dynamic background interference exist in the detection results. In order to speed up ghost removal and suppress the interference of dynamic background, an improved visual background extraction algorithm is proposed. In this method, secondary judgment is added to eliminate ghost pixel interference in the process of spatial transmission of pixels, and the flicker degree of pixels is analyzed to suppress the interference of dynamic background pixels. In order to improve the detection effect, the edge of moving object is obtained by edge detection method, then filled and fused with the detected object. Finally, the detection of moving object is optimized by means of median filtering and mathematical morphology. The simulation results show that the improved algorithm accelerate ghost removal, effectively suppresses the noise interference caused by dynamic background, and improves the accuracy of moving object detection.

关键词： moving object detection Visual background extraction Ghost removal Dynamic background Edge detection

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：