文本检测最近文章检索（1）

期刊

1）Evaluating and improving lexical resources for detecting signs of depression in text
评价和改进词汇资源，用于检测文本中的抑郁迹象
作者： David E. Losada, Pablo Gamallo
作者单位： 1grid.11794.3a, 0000000109410645, Centro Singular de Investigación en Tecnoloxías da Información (CiTIUS), Universidade de Santiago de Compostela, Santiago de Compostela, Spain
刊名： Language Resources and Evaluation, 2020, Vol.54 (6), pp.1-24
来源数据库： Springer Nature Journal
DOI： 10.1007/s10579-018-9423-1
关键词： Depression screening; Depression lexicon; Lexicon evaluation; Lexicon expansion; Text analysis; Natural language processing;
英文摘要： Abstract(#br)While considerable attention has been given to the analysis of texts written by depressed individuals, few studies were interested in evaluating and improving lexical resources for supporting the detection of signs of depression in text.In this paper, we present a search-based methodology to evaluate existing depression lexica. To meet this aim, we exploit existing resources for depression and language use and we analyze which elements of the lexicon are the most effective at revealing depression symptoms. Furthermore, we propose innovative expansion strategies able to further enhance the quality of the lexica.
摘要（#br）尽管人们对抑郁者撰写的文本的分析给予了极大关注，但很少有研究对评估和改进词汇资源以支持文本中抑郁症状的检测感兴趣。基础的方法来评估现有的抑郁症词汇。为了实现这一目标，我们利用现有的抑郁症和语言资源，并分析了词典中哪些元素最有效地揭示了抑郁症的症状。此外，我们提出了能够进一步提高词典质量的创新性扩展策略。
2）Scale robust deep oriented-text detection network面向大规模深度文本检测网络
Zheng Yuqiang; Xie Yuan; Qu Yanyun; Yang Xiaodong; Li Cuihua; Zhang Yan;
JOURNAL:Pattern Recognition
SOURCE:外文期刊
DOI:10.1016/j.patcog.2019.107180
YEAR:2020
PUBLISHER:Pergamon
3）Delaunay triangulation based text detection from multi-view images of natural scene从自然场景的多视图图像中基于Delaunay三角剖分的文本检测
Roy Soumyadip; Shivakumara Palaiahnakote; Pal Umapada; Lu Tong; Kumar Govindaraj Hemantha;
JOURNAL:Pattern Recognition Letters
SOURCE:外文期刊
DOI:10.1016/j.patrec.2019.11.021
YEAR:2020
PAGES:92-100
PUBLISHER:North-Holland
4）Unusual customer response identification and visualization based on text mining and anomaly detection基于文本挖掘和异常检测的异常客户响应识别和可视化
Seo Seungwan; Seo Deokseong; Jang Myeongjun; Jeong Jaeyun; Kang Pilsung;
JOURNAL:Expert Systems with Applications
SOURCE:外文期刊
DOI:10.1016/j.eswa.2019.113111
YEAR:2020
PUBLISER:Pergamon
5）A quadrilateral scene text detector with two-stage network architecture
具有两阶段网络架构的四边形场景文本检测器
Siwei Wang; Yudong Liu; Zheqi He; Yongtao Wang; Zhi Tang;
Wangxuan Institute of Computer Technology; Peking University; No.128 North Zhongguancun Rd; Beijing 100871; China;
ABSTRACT:Abstract(#br)Many of the state-of-the-art methods can only localize scene texts with rotated rectangle boundaries, which may result in incorrect rectification of the detected scene texts and erroneous elimination of proposals or detections during non-maximum suppression (NMS). A few existing methods that can detect scene texts with quadrilateral boundaries, are just based on one-stage architectures or sliding windows scanning and thus have sub-optimal performance. To address these problems, we p… 更多
KEYWORDS:Scene text detection; Deep learning; Quadrilateral regression;
JOURNAL:Pattern Recognition
SOURCE:外文期刊
DOI:10.1016/j.pacog.2020.107230
YEAR:2020
PUBLISHER:Elsevier Ltd
6）Nanotechnology - Quantum Dots; Studies from University of New South Wales Provide New Data on Quantum Dots (A Database for Urdu Text Detection and Recognition in Natural Scene Images)
纳米技术-量子点；新南威尔士大学的研究提供了有关量子点的新数据（用于自然场景图像中乌尔都语文本检测和识别的数据库）
JOURNAL:Physics Week
SOURCE:外文期刊
YEAR:2020
7）TEDLESS – Text detection using least-square SVM from natural scene
Leena Mary Francis; N. Sreenath;
TEDLESS –使用自然场景中的最小二乘SVM进行文本检测
Department of Computer Science and Engineering; Pondicherry Engineering College; Pondicherry 605014; India;
ABSTRACT:Abstract(#br)Text detection from the natural scene is considered to be a challenging problem due to the complex background, varied light intensity at different locations, a large variety of colors, diverse font style and size. This paper focusses on detecting candidate text objects from the scene. The image is initially preprocessed to remove the noise and enhance the contrast. Then the various objects of the scene are marked and extracted forming a pool of objects. A set of candidate text objec… 更多
KEYWORDS:Text detection; Support Vector Machine; Least Square Support Vector Machine; Machine Learning; Natural scene text extraction;
JOURNAL:Journal of King Saud University - Computer and Information Sciences
SOURCE:外文期刊
DOI:10.1016/j.jksuci.2017.09.001
YEAR:2020
PAGES:287-299
PUBLISHER:Elsevier B.V.
8）Unusual customer response identification and visualization based on text mining and anomaly detection基于文本挖掘和异常检测的异常客户响应识别和可视化
Seungwan Seo; Deokseong Seo; Myeongjun Jang; Jaeyun Jeong; Pilsung Kang;
School of Industrial Management Engineering, Korea University, Seoul, South Korea; KEPCO, Data Science Lab, South Korea; Data Machine Intelligence Group, AI center, SK telecom, South Korea;
ABSTRACT:Abstract(#br)The Vehicle Dependability Study (VDS) is a survey study on customer satisfaction for vehicles that have been sold for three years. VDS data analytics plays an important role in the vehicle development process because it can contribute to enhancing the brand image and sales of an automobile company by properly reflecting customer requirements retrieved from the analysis results when developing the vehicle’s next model. Conventional approaches to analyzing the voice of customers (VOC)… 更多
车辆可靠性研究（VDS）是一项针对已售出三年的汽车的客户满意度的调查研究。VDS数据分析在车辆开发过程中起着重要作用，因为它可以通过在开发车辆的下一个模型时正确反映从分析结果中检索到的客户需求，从而有助于提高汽车公司的品牌形象和销量。分析客户声音（VOC）的常规方法… 更多
KEYWORDS:Voice of customers; Keyword network; Local outlier factor; TF-IDF;
JOURNAL:Expert Systems With Applications
SOURCE:外文期刊
DOI:10.1016/j.eswa.2019.113111
YEAR:2020
PUBLISHER:Elsevier Ltd
9）
IOS-Net: An inside-to-outside supervision network for scale robust text detection in the wildIOS-Net：一个内部到外部的监督网络，可以在野外进行规模强大的文本检测
Yuanqiang Cai; Weiqiang Wang; Yuting Chen; Qixiang Ye;
School of Computer and Control Engineering, University of Chinese Academy of Sciences, Beijing, China; University of Michigan, Ann Arbor, United States; School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing, China;
ABSTRACT:Abstract(#br)Accurately detecting scene text is a challenging task due to perspective distortion, scale variance, varied orientations, uneven illumination. Among them, scale variance has always been a core issue and generally involves two types: various size and diverse aspect ratios of the text regions. In contrast to most existing approaches focusing on addressing one type of scale variance, this paper presents a novel inside-to-outside supervision network (IOS-Net) that can well tackle both t… 更多
摘要：由于透视变形，比例尺变化，方向变化，照明不均匀，准确检测场景文本是一项艰巨的任务。其中，比例差异一直是核心问题，通常涉及两种类型：文本区域的各种大小和不同的纵横比。与此相反，以重点解决一类规模差异的大部分现有的方法，可以很好解决本文提出了一种新型的由内到外的监督网络（IOS-网）两者的… 更多
KEYWORDS:Text detection; Various sizes; Diverse aspect ratios; Inside-to-outside supervision; Position-sensitive segmentation;
JOURNAL:Pattern Recognition
SOURCE:外文期刊
DOI:10.1016/j.patcog.2020.107304
YEAR:2020
PUBLISHER:Elsevier Ltd
10）Review of Scene Text Detection and Recognition
场景文本检测与识别综述
作者： Han Lin, Peng Yang, Fanlong Zhang
作者单位： 1School of Information Engineering, Nanjing Audit University, 211815, Jiangshu, China
2School of Information Engineering, Nanchang Hangkong University, 330063, Jiangxi, China
刊名： Archives of Computational Methods in Engineering: State of the Art Reviews, 2020, Vol.27 (2), pp.433-454
来源数据库： Springer Nature Journal
DOI： 10.1007/s11831-019-09315-1
英文摘要： Abstract(#br)Scene texts contain rich semantic information which may be used in many vision-based applications, and consequently detecting and recognizing scene texts have received increasing attention in recent years. In this paper, we first introduce the history and progress of scene text detection and recognition, and classify conventional methods in detail and point out their advantages as well as disadvantages. After that, we study these methods and illustrate the corresponding key issues and techniques, including loss function, multi-orientation, language model and sequence labeling. Finally, we describe commonly used benchmark datasets and evaluation protocols, based on which the performance of representative scene text detection and recognition methods are analyzed and compared.
摘要（#br）场景文本包含丰富的语义信息，可在许多基于视觉的应用程序中使用，因此，近年来，检测和识别场景文本已受到越来越多的关注。在本文中，我们首先介绍了场景文本检测和识别的历史和进展，并对传统方法进行了详细分类，并指出了它们的优缺点。之后，我们将研究这些方法并说明相应的关键问题和技术，包括损失函数，多方向，语言模型和序列标记。最后，我们描述了常用的基准数据集和评估协议，在此基础上，对代表性场景文本检测和识别方法的性能进行了分析和比较。
11）一种面向结构化文本图像识别的深度学习模型
唐三立程战战钮毅雷鸣
杭州海康威视数字技术股份有限公司
摘要：提出一种面向结构化文本识别的端到端训练的深度学习模型,旨在使用一个兼具文本检测、识别与信息结构化能力的深度学习模型对包含结构化文本的图像进行文本信息提取,从而高效且精确地获得图像中各文本字段的位置、内容及属性。实验表明:相比于现有方案,提出的结构化文本图像识别模型能提高结构化文本识别效率与性能,同时减少模型个数和模型参数量,提高开发效率。
基金：国家重点研发计划资助项目(2018YFC0807706)；
关键词：结构化文本识别; 文本检测; 文本识别; 信息结构化;
DOI：10.13954/j.cnki.hdu.2020.02.008
分类号：TP18;TP391.41
12）A node-priority based large-scale overlapping community detection using evolutionary multi-objective optimization
基于节点优先级的进化多目标大规模重叠社区检测
作者： Zhengyi Chai, Shijiao Liang
作者单位： 1School of Computer Science and Technology, Tianjin Polytechnic University, 300387, Tianjin, China
刊名： Evolutionary Intelligence, 2020, Vol.13 (6684), pp.59-68
来源数据库： Springer Nature Journal
DOI： 10.1007/s12065-019-00250-5
关键词： Community detection; Multi-objective optimization; Large-scale network; Pareto fronts; Node priority;
英文摘要： Abstract(#br)Community structure is one of the most important features in complex networks. However, with increasing of network scale, some existing methods cannot effectively detect the community structure of complex network, and the available methods mostly aimed at non-overlapping networks. In this paper, we focus on overlapping community detection in large-scale networks, because most of the communities in real-world networks are overlapped. In order to improve the accuracy of large-scale overlapping community detection, we suggest a community detection method based on node priority. The proposed algorithm has two advantages: (1) We define a priority function f NN…
摘要（#br）社区结构是复杂网络中最重要的功能之一。但是，随着网络规模的扩大，一些现有的方法无法有效地检测复杂网络的社区结构，而可用的方法主要针对非重叠网络。在本文中，我们将重点放在大型网络中的重叠社区检测上，因为现实世界网络中的大多数社区都是重叠的。为了提高大规模重叠社区检测的准确性，提出了一种基于节点优先级的社区检测方法。该算法具有两个优点：（1）我们定义了一个优先级函数f NN …
13）Research on Privacy Disclosure Detection Method in Social Networks Based on Multi-Dimensional Deep Learning
基于多维深度学习的社交网络隐私披露检测方法研究
作者:
Yabin Xu,12,Xuyang Meng,Yangyang Li,Xiaowei Xu,4
作者背景:N/A
DOI:10.32604/cmc.2020.05825
文章关键词:Social networks,Privacy disclosure detection,Multi-dimensional features,Text classification,Convolutional neural network.
原文摘要:In order to effectively detect the privacy that may be leaked through socialnetworks and avoid unnecessary harm to users, this paper takes microblog as the researchobject to study the detection of privacy disclosure in social networks. First, we performfast privacy leak detection on the currently published text based on the fastText model. Inthe case that the text to be published contains certain private information, we fullyconsider the aggregation effect of the private information leaked by different channels,and establish a convolution neural network model based on multi-dimensional features(MF-CNN to detect privacy disclosure comprehensively and accurately. Theexperimental results show that the proposed method has a higher accuracy of privacydisclosure detection and can meet the real-time requirements of detection.
为了有效检测社交网络中可能泄露的隐私并避免对用户造成不必要的伤害，本文以微博为研究对象，研究社交网络中隐私披露的检测。首先，我们基于fastText模型对当前发布的文本执行快速的隐私泄漏检测。在要发表的文本包含某些私人信息的情况下，我们充分考虑不同渠道泄漏的私人信息的聚集效应，并建立基于多维特征的卷积神经网络模型（MF-CNN），以全面检测隐私泄露；实验结果表明，该方法具有较高的隐私公开检测精度，能够满足实时检测要求。
14）Multimedia Computing; Findings on Multimedia Computing Detailed by Investigators at University of Science and Technology China (Ab-lstm: Attention-based Bidirectional Lstm Model for Scene Text Detection)
多媒体计算；中国科学技术大学的研究人员详细介绍了多媒体计算的发现（Ab-lstm：基于注意力的双向Lstm模型用于场景文本检测）
JOURNAL:Computer Technology Journal
SOURCE:外文期刊
YEAR:2020
15）Multi-level Fuzzy Based Renyi Entropy for Linguistic Classification of Texts in Natural Scene Images
基于多级模糊的Renyi熵用于自然场景图像文本的语言分类
作者： Angia Venkatesan Karpagam, Mohan Manikandan
作者单位： 1grid.252262.3, 0000 0001 0613 6919, Department of Electronics Engineering, Madras Institute of Technology, Anna University Chennai, Chennai, India
刊名： International Journal of Fuzzy Systems, 2020, Vol.22 (4), pp.438-449
来源数据库： Springer Nature Journal
DOI： 10.1007/s40815-019-00654-6
关键词： Text detection; Linguistic classification; CIE-Lab color space; Stroke width transform; Hu moments; Support vector machine;
英文摘要： Abstract(#br)This paper focuses on linguistic classification of scene texts in natural scene images. In this paper, an attempt is made to localize texts based on multi-level thresholding by fuzzy-based Renyi entropy. Complex natural scene images with diversified challenges are considered. A set of heuristic rules comprising geometric filters and stroke width transform govern the process of locating potential text regions. The scene images may contain more than one language, where text recognition by optical character recognition system becomes challenging. Manual intervention is needed to specify the language of each text. To overcome this hurdle, linguistic classification of text regions is suggested in this paper. The proposed method is validated using publicly available dataset—MSRA-TD500. Results show that fuzzy-based Renyi entropy thresholding is able to segment the foreground text from complex natural scene images. Geometric filters could capture the inherent uniformity of the text. Stroke width transform eliminates the non-text regions. The performance measures such as precision, recall and F -measures are 78%, 77% and 76%, respectively. This shows the ability of the algorithm to extract the text from the scenes. The geometric feature such as area and corner shows better variation in discriminating the linguistic texts. Further, the first three Hu moment features also contribute remarkable role in analyzing the shape of extracted text regions. The classifier based on support vector machine (SVM) yields classification accuracy of 85.45% in discriminating English and Chinese alphabets. Area under the ROC curve (AUC) is 0.851 for SVM classifier. The proposed methodology has proved its robustness against common degradations, such as uneven illumination, varying font characteristics and blurring effects. Experimental results show that our method achieves better performance in linguistic classification.
摘要（#br）本文重点研究自然场景图像中场景文本的语言分类。本文尝试通过基于模糊的Renyi熵的多级阈值技术对文本进行本地化。考虑具有多种挑战的复杂自然场景图像。一组包括几何过滤器和笔触宽度变换的启发式规则控制着定位潜在文本区域的过程。场景图像可能包含多种语言，通过光学字符识别系统进行文本识别变得很有挑战性。需要人工干预以指定每个文本的语言。为了克服这一障碍，本文提出了文本区域的语言分类。使用公开可用的方法验证了所提出的方法数据集-MSRA-TD500。结果表明，基于模糊的人意熵阈值能够分割复杂自然场景图像中的前景文本。几何过滤器可以捕获文本的固有一致性。笔划宽度变换消除了非文本区域。诸如精度，召回率和F度量之类的性能度量分别为78％，77％和76％。这显示了算法从场景中提取文本的能力。诸如面积和角之类的几何特征在区分语言文本时显示出更好的变化。此外，前三个Hu矩特征在分析提取的文本区域的形状方面也发挥了重要作用。基于支持向量机（SVM）的分类器在区分英文和中文字母时的分类精度为85.45％。对于SVM分类器，ROC曲线下的面积（AUC）为0.851。所提出的方法论证明了其对于常见降级（例如照明不均匀，字体特征变化和模糊效果）的鲁棒性。实验结果表明，该方法在语言分类中取得了较好的效果。
16）Total-Text: toward orientation robustness in scene text detection
Total-Text：场景文本检测中的方向稳健性
作者： Chee-Kheng Ch’ng, Chee Seng Chan, Cheng-Lin Liu
刊名： International Journal on Document Analysis and Recognition (IJDAR, 2020, Vol.23 (1), pp.31-52
来源数据库： Springer Nature Journal
DOI： 10.1007/s10032-019-00334-z
17）Method for Detecting Chinese Texts in Natural Scenes Based on Improved Faster R-CNN
基于改进的快速R-CNN的自然场景中文文本检测方法
作者:
Shuhua Liu,Hua Ban,Yu Song,Mengyu Zhang,Fengqin Yang
作者背景:N/A
DOI:10.1142/S021800142053002X
文章关键词:Faster RCNN,Inception ResNet,Text detection
原文摘要:In this study, a natural scene text detection method based on the improved faster region-based convolutional neural network (R-CNN) is proposed. This method extracts image features with the Inception-ResNet architecture, adopts a region proposal network to generate region proposals for the extracted features, merges the fine-tuned features with the region proposals, and finally, uses Fast R-CNN to classify and locate text. The proposed method solves the problems of varying text sizes and the text being obscured in the image. Compared with the original Faster R-CNN, the multilevel Inception-ResNet network model presented in this study can extract deeper text features. The extracted feature map is further sparsely represented by Reduction B, Inception ResNet C and Avg Pool, and then is fused with text regions obtained by the text feature mapping lower layer network to acquire the exact text regions. The text detection method presented in this study is tested on the 2017 dataset of ICDAR2017 Competition on Reading Chinese Text in the Wild (RCTW-17), which contains a large number of distorted, blurry, different scale and size texts. An accuracy of 76.4% is achieved in this platform, thereby proving the efficiency of the proposed method.
本文提出了一种基于改进的基于快速区域卷积神经网络（R-CNN）的自然场景文本检测方法。该方法采用Inception-ResNet架构提取图像特征，采用区域提议网络为提取的特征生成区域提议，将微调的特征与区域提议合并，最后使用Fast R-CNN进行文本分类和定位。所提出的方法解决了文本尺寸变化以及图像中文本被遮盖的问题。与原始的Faster R-CNN相比，本研究中提出的多层Inception-ResNet网络模型可以提取更深的文本特征。提取的特征图进一步由还原B，Inception ResNet C和Avg Pool稀疏表示，然后与由文本特征映射下层网络获得的文本区域融合，以获取确切的文本区域。本研究中介绍的文本检测方法已在ICDAR2017野外阅读中文比赛2017（RCTW-17）的数据集上进行了测试，其中包含大量失真，模糊，不同比例和大小的文本。在该平台上实现了76.4％的精度，从而证明了所提出方法的效率。
18）Towards automated detection of psychosocial risk factors with text mining.
旨在通过文本挖掘自动检测社会心理风险因素
Uronen L; Moen H; Teperi S; Martimo K-P; Hartiala J; Salanterä S;
Department of Nursing Science, University of Turku, Turku, Finland.; Department of Future Technologies, University of Turku, Turku, Finland.; Department of Mathematics and Statistics, University of Turku, Turku, Finland.; Ilmarinen Mutual Pension Insurance Company, Helsinki, Finland.; Department of Medicine, University of Turku, Turku, Finland.;
ABSTRACT:BACKGROUND(#br)Psychosocial risk factors influence early retirement and absence from work. Health checks by occupational health nurses (OHNs) may prevent deterioration of work ability. Health checks are documented electronically mostly as free text, and therefore the effect of psychological risk factors on working capacity is difficult to detect.(#br)AIMS(#br)To evaluate the potential of text mining for automated early detection of psychosocial risk factors by examining health check free-text do… 更多
KEYWORDS:Health check; Occupational health; Psychosocial risk factors; Text mining;
JOURNAL:Occupational medicine (Oxford, England)
SOURCE:外文期刊
DOI:10.1093/occmed/kqaa022
YEAR:2020
PUBLISHER:Pubmed
19）多视图场景中的文本共检测。
Text Co-detection in Multi-view Scene.
Wang Chuan; Fu Huazhu; Yang Liang; Cao Xiaochun;
Wang Chuan_; Fu Huazhu_; Yang Liang_; Cao Xiaochun_;
ABSTRACT:Multi-view scene analysis has been widely explored in computer vision, including numerous practical applications. The texts in multi-view scenes are often detected by following the existing text detection method in a single image, which however ignores the multi-view corresponding constraint. The multi-view correspondences may contain structure, location information and assist difficulties induced by factors like occlusion and perspective distortion, which are deficient in the single image scene… 更多
摘要：在计算机视觉中已经广泛探索了多视图场景分析，包括许多实际应用。通常通过遵循单个图像中的现有文本检测方法来检测多视图场景中的文本，但是该方法忽略了多视图对应的约束。多视图对应关系可能包含结构，位置信息，并会因遮挡和透视变形等因素而导致的困难，而这些困难在单个图像场景中是不足的… 更多
JOURNAL:IEEE transactions on image processing : a publication of the IEEE Signal Processing Society
SOURCE:外文期刊
DOI:10.1109/TIP.2020.2973511
YEAR:2020
PUBLISHER:Pubmed
20）Datalogic USA Inc.; Patent Issued for Robust String Text Detection For Industrial Optical Character Recognition (USPTO 10,552,699)
Datalogic USA Inc .；专利权，用于工业光学字符识别的健壮字符串文本检测（USPTO 10,552,699）
JOURNAL:Journal of Robotics & Machine Learning
SOURCE:外文期刊
YEAR:2020
21）Using the “Who, What, and When” of free text documentation to improve hospital infectious disease surveillance
使用“谁，什么，什么时候”的免费文本文档来改善医院传染病的监测
Philip Zachariah; Alexandra Hill-Ricciuti; Lisa Saiman; Karthik Natarajan;
Department of Pediatrics, Columbia University Irving Medical Center, New York, NY; Department of Infection Prevention & Control, NewYork-Presbyterian Hospital, New York, NY; Department of Biomedical Informatics, Columbia University Irving Medical Center, New York, NY;
ABSTRACT:We demonstrate a novel method of using unstructured health data for infectious disease surveillance. A model incorporating the dynamics of documentation of a test diagnosis (UTI) in free text, without using grammatical or syntactic analysis, achieved performance comparable to ICD-10 codes (sensitivity 57.3, positive predictive value 69.5%, negative predictive value 95.9%) and detected missed cases (15% of total).
KEYWORDS:Detection; Electronic;
JOURNAL:AJIC: American Journal of Infection Control
SOURCE:外文期刊
DOI:10.1016/j.ajic.2020.01.001
YEAR:2020
PUBLISHER:Elsevier Inc.
22）
Datalogic USA Inc.; Patent Issued for Robust String Text Detection For Industrial Optical Character Recognition (USPTO 10,552,699)
Datalogic USA Inc .；专利权，用于工业光学字符识别的健壮字符串文本检测（USPTO 10,552,699）
JOURNAL:Journal of Robotics & Machine Learning
SOURCE:外文期刊
YEAR:2020
23）工程; 北京交通大学研究人员的工程研究成果（采用深度学习方法对医学检验报告的图像进行文本检测和识别）
Engineering; Researchers from Beijing Jiaotong University Report Findings in Engineering (Text Detection and Recognition for Images of Medical Laboratory Reports With a Deep Learning Approach
JOURNAL:Journal of Engineering
SOURCE:外文期刊
YEAR:2020
PUBLISHER:NewsRx
24）Bottom-Up Scene Text Detection with Markov Clustering Networks
采用马尔可夫聚类网络进行自下而上的场景文本检测
作者： Zichuan Liu, Guosheng Lin, Wang Ling Goh
刊名： International Journal of Computer Vision, 2020(prepublish), pp.1-24
来源数据库： Springer Nature Journal
DOI： 10.1007/s11263-020-01298-y
25）Realtime multi-scale scene text detection with s cale-based r egion p roposal n etwork使用基于场景的区域网络实时多尺度场景文本检测
Wenhao He; Xu-Yao Zhang; Fei Yin; Zhenbo Luo; Jean-Marc Ogier; Cheng-Lin Liu;
National Laboratory of Pattern Recognition (NLPR), Institute of Automation of Chinese Academy of Sciences, Beijing 100190, China; University of Chinese Academy of Sciences (UCAS), Beijing 100049, China; Beijing Samsung Telecom R&D Center, Beijing, China; L3i Laboratory, University of La Rochelle, La Rochelle, France; CAS Center for Excellence of Brain Science and Intelligence Technology, Beijing 100190, China;
ABSTRACT:Abstract(#br)Multi-scale approaches have been widely used for achieving high accuracy for scene text detection, but they usually slow down the speed of the whole system. In this paper, we propose a two-stage framework for realtime multi-scale scene text detection. The first stage employs a novel S cale-based R egion P roposal N etwork (SRPN) which can localize text of wide scale range and estimate text scale efficiently. Based on SRPN, non-text regions are filtered out, and text region proposals… 更多
摘要：多尺度方法已广泛用于实现场景文本检测的高精度，但通常会减慢整个系统的速度。在本文中，我们提出了一个用于实时多尺度场景文本检测的两阶段框架。第一阶段采用一种新颖的基于S级的区域计划网络（SRPN），它可以定位范围广泛的文本并有效地估计文本规模。基于SRPN，非文本区域被过滤掉，文本区域建议… 更多
KEYWORDS:Scene text detection; Multi-scale; Speedup; Scale-based region proposal network;
JOURNAL:Pattern Recognition
SOURCE:外文期刊
DOI:10.1016/j.patcog.2019.107026
YEAR:2020
PUBLISHER:Elsevier Ltd
26）Semantic similarity and text summarization based novelty detection
基于语义相似度和文本摘要的新颖性检测
作者： Sushil Kumar, Komal Kumar Bhatia
作者单位： 1Department of Computer Engineering, J.C. Bose University of Science and Technology, YMCA, 121006, Faridabad, Haryana, India
刊名： SN Applied Sciences, 2020, Vol.2 (6), pp.1-10
来源数据库： Springer Nature Journal
DOI： 10.1007/s42452-020-2082-z
关键词： Dice coefficient; Hash value; Novelty detection; Semantic similarity; Word net 3.0;
英文摘要： Abstract(#br)Current web crawlers search the queries at very high speed, but the problem of novelty detection or redundant information still persists. It consumes precious time and memory of users in search of the new document over the internet. In this paper, an innovative novelty detection mechanism is proposed, which can be appended with the current web crawlers. The proposed mechanism first summarizes the text, based on ontology, and from the obtained summary, semantic similarity is calculated using word net 3.0. The hash value is then calculated using the winnowing algorithm. This hash value of the document is matched with others using the Dice coefficient to calculate the similarity index. Based on the threshold chosen for similarity, the document is treated either as novel or not…
摘要（#br）当前的网络爬虫以很高的速度搜索查询，但是新颖性检测或冗余信息的问题仍然存在。在Internet上搜索新文档会浪费用户宝贵的时间和内存。本文提出了一种新颖的新颖性检测机制，该机制可以与当前的网络爬虫一起使用。所提出的机制首先基于本体对文本进行摘要，然后从所获得的摘要中使用词网3.0计算语义相似度。然后使用风选算法计算哈希值。使用Dice系数将文档的哈希值与其他哈希值进行匹配，以计算相似性指数。根据为相似性选择的阈值，将文档视为是新文档还是非新文档。…
27）Selective feature connection mechanism: Concatenating multi-layer CNN features with a feature selector选择性要素连接机制：将多层CNN要素与要素选择器串联
Chen Du; Yanna Wang; Chunheng Wang; Cunzhao Shi; Baihua Xiao;
State Key Laboratory of Management and Control for Complex Systems, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China; University of Chinese Academy of Sciences, Beijing, China;
ABSTRACT:Abstract(#br)Different layers of deep convolutional neural networks(CNNs) can encode different-level information. High-layer features always contain more semantic information, and low-layer features contain more detail information. However, low-layer features suffer from the background clutter and semantic ambiguity. During visual recognition, the feature combination of the low-layer and high-level features plays an important role in context modulation. If directly combining the high-layer and l… 更多
摘要：深度卷积神经网络（CNN）的不同层可以编码不同级别的信息。高层功能始终包含更多的语义信息，而低层功能则包含更多的详细信息。然而，低层特征遭受背景混乱和语义歧义的困扰。在视觉识别期间，低层和高级特征的特征组合在上下文调制中起重要作用。如果直接将高层和底层结合起来，… 更多
KEYWORDS:Feature combination; Network architecture; Selective feature connection mechanism; Convolutional neural network;
JOURNAL:Pattern Recognition Letters
SOURCE:外文期刊
DOI:10.1016/j.patrec.2019.11.015
YEAR:2020
PAGES:108-114
PUBLISHER:Elsevier B.V.
28）Delaunay triangulation based text detection from multi-view images of natural scene从自然场景的多视图图像中基于Delaunay三角剖分的文本检测
Soumyadip Roy; Palaiahnakote Shivakumara; Umapada Pal; Tong Lu; Govindaraj Hemantha Kumar;
Computer Science And Engineering, Heritage Institute Of Technology, Kolkata, India; Faculty of Computer Science and Information Technology, University of Malaya, Kuala Lampur, Malaysia; Computer Vision and Pattern Recognition Unit, Indian Statistical Institute, Kolkata, India; National Key Lab of Novel Software Technology, Nanjing University, Nanjing, China; Department of Studies in Computer Science, University of Mysore, Karnataka, India;
ABSTRACT:Abstract(#br)Text detection in the wild is still considered as a challenging issue to the researchers because of its several real time applications like forensic application, where CCTV camera captures images at different angles of the same scene. Unlike the existing methods that consider a single view captured orthogonally for text detection, this paper considers multi-view (view-1 and view-2 of the same spot) of the same scene captured at different angles or different height distances for text… 更多
摘要：摘要（#br）野外文本检测仍被认为是研究人员面临的挑战性问题，因为它具有多种实时应用，例如取证应用，其中CCTV摄像机在同一场景的不同角度捕获图像。与现有的考虑将单个视图正交捕获以进行文本检测的方法不同，本文考虑了以不同角度或不同高度距离捕获的同一场景的多视图（同一点的view-1和view-2）用于文本检测。对于同一场景的每一对，所提出的方法基于Delaunay三角剖分（DT）提取描述文本成分特征的特征，即DT的角点，面积和空腔。通过余弦距离测量比较视图1和视图2中相应DT的特征，以估计相应视图1和视图2的两个组件之间的相似性。如果该对满足相似条件，则将这些组件视为候选文本组件（CTC）。换句话说，它们是满足相似条件的view-1和view-2的通用组件。该方法从视图1和视图2的每个CTC中，使用卡方和余弦距离测度，基于CTC和邻居分量之间的相似程度估计，找到最接近的邻居分量以恢复同一文本行的分量。此外，所提出的方法使用识别步骤，通过比较视图1和视图2的识别结果来检测正确的文本。相同的识别步骤用于消除误报，以提高所提出方法的性能。在我们自己的数据集上的实验结果包含了不同情况的图像对，而标准数据集即ICDAR 2013，MSRATD-500，CTW1500，Total-text，ICDAR 2017 MLT和COCO-text，表明该方法的性能优于现有方法。 还原
JOURNAL:Pattern Recognition Letters
SOURCE:外文期刊
DOI:10.1016/j.patrec.2019.11.021
YEAR:2020
PAGES:92-100
PUBLISHER:Elsevier B.V.
29）
Robustly detect different types of text in videos
稳健地检测视频中不同类型的文本
作者： Yuanqiang Cai, Weiqiang Wang
刊名： Neural Computing and Applications, 2020(prepublish), pp.1-14
来源数据库： Springer Nature Journal
DOI： 10.1007/s00521-020-04729-6
30）基于MobileNet模型的钢材表面字符检测识别算法
艾梦琴陶青川
四川大学电子信息学院
摘要：鉴于采集钢材板坯号图像的现场环境十分恶劣,因此通常难以采集到高清图像,并且采集图像目标区域会出现模糊、扭曲和倾斜等显现,对传统的检测、识别算法造成很大的阻碍,造成目标检测精度、速度不高。为了提高定位、识别板坯号的精度和速度,需要应用计算机视觉技术与深度学习技术,帮助钢厂达到智能制造的要求。研究并实现一种快速定位钢板坯料号目标区域算法并端到端的识别算法,能够达到实时快速的检测、识别。
关键词：钢材表面; 卷积神经网络; 文本检测; 端到端识别;
分类号：TP391.41;TG142.1
31）基于神经网络的自然场景方向文本检测器
周铂焱杨鹏
南昌航空大学信息工程学院
摘要：场景文本检测是场景文本识别中重要的一步,也是一个具有挑战性的问题。不同于一般的目标检测,场景文本检测的主要挑战在于自然场景图像中的文本具有任意方向,小的尺寸,以及多种宽高比。论文在TextBoxes[8]的基础上进行改进,提出了一个适用于任意方向文本的检测器,命名为OSTD（Oriented Scene Text Detector）,可以有效且准确地检测自然场景中任意方向的文本。论文在公共数据集上对提出OSTD的进行评估。所有实验结果都表明,无论在准确性,还是实时性方面OSTD都是极具竞争力的方法。在1024×1024的ICDAR2015 Incidental Text数据集[16]上,OSTD的F-Measure=0.794,FPS=10.7。
基金：国家自然科学基金项目（编号：61662048,61363050）资助；
关键词：自然场景图像; 文本检测; 神经网络;
分类号：TP391.41;TP183
32）一种基于文本检测的书脊定位方法
崔晨任明武
南京理工大学计算机科学与工程学院
摘要：书脊定位是实现图书管理自动化的重要技术,通过对定位分割出的书脊图像进行图像匹配或文本识别获得图书信息,可大大减小图书检索、整理的人力劳动。论文提出了一种基于文本检测的书脊区域粗选方法,首先通过序贯分割算法检测图像中的字符整体区域,然后根据字符宽度和距离将同属于一本书的字符加入相似字符集合,根据集合内的字符中心和字符宽度计算候选书脊区域,最后通过支持向量机分类器精选书脊区域。相比于已有的书脊定位方法,论文算法在光照敏感、相邻书脊颜色对比度敏感、书脊多角度倾斜检测等方面进行了改善,在实验中取得了较好的定位成功率。
关键词：书脊定位; 区域合并; 文本检测; HOG; SVM;
分类号：TP181;TP391.41
33）Active contour-based detection of estuarine dolphin whistles in spectrogram images基于主动轮廓的光谱图中的河口海豚哨声检测
O.M. Serra; F.P.R. Martins; L.R. Padovese;
Universidade de São Paulo; Escola Politécnica; Department of Mechanical Engineering; Av. Prof. Mello Moraes; 2231; 05508-030 São Paulo; Brazil;
ABSTRACT:Abstract(#br)An algorithm for detecting tonal vocalizations from estuarine dolphin ( Sotalia guianensis ) specimens without interference of a human operator was developed. The raw audio data collected from a passive monitoring sensor in the Cananéia underwater soundscape was converted to spectrogram images. Detection is a four-step method: first, ridge maps are obtained from the spectrogram images; second, a probabilistic Hough transform algorithm is applied to detect ridges similar to thick lin… 更多
摘要：摘要（#br）提出了一种在没有人工干预的情况下，从河口海豚（Sotalia guianensis）标本中检测发声的算法。从Cananéia水下音景中的无源监视传感器收集的原始音频数据被转换为频谱图图像。检测是四个步骤的方法：首先，从频谱图图像中获取山脊图；然后，其次，应用概率霍夫变换算法来检测类似于粗线段的脊，称为线状，通过主动轮廓算法将其调整为图像中哨声的几何形状。第三，从每个检测到的口哨的几何形状构建特征向量，具有9个描述性特征；第四，将检测结果馈送到随机森林分类器，以通过检测过程解析错误。我们开发了一种系统，可以将所检测到的特征模式分类为sian guianensis哨声或随机空检测。我们获得了0.977的准确度和F81的0.981。 还原
KEYWORDS:( Sotalia guianensis; Spectrogram; Hough transform; Active contours; Random forest; Machine learning);
JOURNAL:Ecological Informatics
SOURCE:外文期刊
DOI:10.1016/j.ecoinf.2019.101036
YEAR:2020
PUBLISHER:Elsevier B.V.
34）顾及目标关联的自然场景文本检测
易尧华何婧婧卢利琼汤梓伟
武汉大学印刷与包装系
摘要：目的目前基于卷积神经网络(CNN)的文本检测方法对自然场景中小尺度文本的定位非常困难。但自然场景图像中文本目标与其他目标存在很强的关联性,即自然场景中的文本通常伴随特定物体如广告牌、路牌等同时出现,基于此本文提出了一种顾及目标关联的级联CNN自然场景文本检测方法。方法首先利用CNN检测文本目标及包含文本的关联物体目标,得到文本候选框及包含文本的关联物体候选框;再扩大包含文本的关联物体候选框区域,并从原始图像中裁剪,然后以该裁剪图像作为CNN的输入再精确检测文本候选框;最后采用非极大值抑制方法融合上述两步生成的文本候选框,得到文本检测结果。结果本文方法能够有效地检测小尺度文本,在ICDAR-2013数据集上召回率、准确率和F值分别为0. 817、0. 880和0. 847。结论本文方法顾及自然场景中文本目标与包含文本的物体目标的强关联性,提高了自然场景图像中小尺度文本检测的召回率。
基金：国家科技重大专项基金项目(2017ZX01030102)~~；
关键词：自然场景; 文本检测; 小尺度文本; 目标关联; 级联卷积神经网络;
分类号：TP391.41;TP183
35）Machine Learning; Reports on Machine Learning from University of Michigan Provide New Insights (Detecting Substance-related Problems In Narrative Investigation Summaries of Child Abuse and Neglect Using Text Mining and Machine Learning)
机器学习；来自密歇根大学的机器学习报告提供了新的见解（使用文本挖掘和机器学习检测虐待和忽视儿童的叙事调查摘要中与物质相关的问题）
JOURNAL:Computers, Networks & Communications
SOURCE:外文期刊
YEAR:2020
36）自然场景图片中的文本检测和定位
司飞
中国传媒大学
摘要：本文针对自然场景下背景复杂、字符残缺、阴影遮挡等多种情形,为了提高文本定位的准确率,提出了一种改进的最大稳定极值区域(Maximally Stable External Region,MSER)的方法。首先将图片进行通道分离,对每个颜色空间分别提取MSER区域,然后将每个通道得到的MSER进行融合,再通过神经网络模型进行背景和文本分离。最后将得到的字符进行文本合并,得到最终的文本行,通过试验,此方法在ICDAR2011、ICDAR2013公开数据集上有较好的试验效果。
关键词：卷积神经网络（CNN）; 文本定位; 最大稳定极值区域（MSER）;
分类号：TP391.41
37）SynthText3D: synthesizing scene text images from 3D virtual worlds
SynthText3D：合成来自3D虚拟世界的场景文本图像
作者： Minghui Liao, Boyu Song, Shangbang Long, Minghang He, Cong Yao, Xiang Bai
作者单位： 1School of Electronic Information and Communications, Huazhong University of Science and Technology, 430074, Wuhan, China
2School of Electronics Engineering and Computer Science, Peking University, 100871, Beijing, China
3School of Economics, Peking University, 100871, Beijing, China
4MEG VII, 100080, Beijing, China
刊名： Science China Information Sciences, 2020, Vol.63 (2), pp.220104-3220
来源数据库： Springer Nature Journal
DOI： 10.1007/s11432-019-2737-0
关键词： Optical character recognition (OCR); Synthetic data; Scene text detection; 3D; Deep learning;
英文摘要： Abstract(#br)With the development of deep neural networks, the demand for a significant amount of annotated training data becomes the performance bottlenecks in many fields of research and applications. Image synthesis can generate annotated images automatically and freely, which gains increasing attention recently. In this paper, we propose to synthesize scene text images from the 3D virtual worlds, where the precise descriptions of scenes, editable illumination/visibility, and realistic physics are provided. Different from the previous methods which paste the rendered text on static 2D images, our method can render the 3D virtual scene and text instances as an entirety. In this way, real-world variations, including complex perspective transformations, various illuminations, and…
38）
Energy; GMR Institute of Technology Researchers Provide New Data on Energy (Text detection and recognition in raw image dataset of seven segment digital energy meter display)
能源; GMR技术研究所提供有关能源的新数据（七段数字电表显示的原始图像数据集中的文本检测和识别）
JOURNAL:Energy Weekly News
SOURCE:外文期刊
YEAR:2020
YEAR:2020
39）Scene Text Detection with Polygon Offsetting and Border Augmentation
具有多边形偏移和边界增强的场景文本检测
作者:
Thananop Kobchaisawat,Thanarat H. Chalidabhongse,Shin’ichi Satoh
作者背景:
Department of Computer Engineering, Chulalongkorn University, Bangkok 10330, Thailand; thananop.k@student.chula.ac.th
Center of Excellence in Infrastructure Management, Chulalongkorn University, Bangkok 10330, Thailand
Research Group on Applied Computer Engineering Technology for Medicine and Healthcare, Chulalongkorn University, Bangkok 10330, Thailand
National Institute of Informatics, Tokyo 101-8430, Japan; satoh@nii.ac.jp
展开
DOI:10.3390/electronics9010117
文章关键词:Scene text detection,Curved text detection,Convolutional neural networks
原文摘要:Scene text localization is a very crucial step in the issue of scene text recognition. The major challenges—such as how there are various sizes, shapes, unpredictable orientations, a wide range of colors and styles, occlusion, and local and global illumination variations—make the problem different from generic object detection. Unlike existing scene text localization methods, here we present a segmentation-based text detector which can detect an arbitrary shaped scene text by using polygon offsetting, combined with the border augmentation. This technique better distinguishes contiguous and arbitrary shaped text instances from nearby non-text regions. The quantitative experimental results on public benchmarks, ICDAR2015, ICDAR2017-MLT, ICDAR2019-MLT, and Total-Text datasets demonstrate the performance and robustness of our proposed method, compared to previous approaches which have been proposed.
场景文本本地化是场景文本识别问题中非常关键的一步。主要挑战（例如，如何具有各种大小，形状，不可预测的方向，广泛的颜色和样式，遮挡以及局部和全局照明变化）使该问题不同于普通物体检测。与现有的场景文本本地化方法不同，这里我们介绍一种基于分段的文本检测器，该检测器可以通过使用多边形偏移结合边界增强来检测任意形状的场景文本。此技术可以更好地区分连续的和任意形状的文本实例与附近的非文本区域。在公开基准，ICDAR2015，ICDAR2017-MLT，ICDAR2019-MLT和Total-Text数据集上的定量实验结果证明了我们提出的方法的性能和鲁棒性，
40）多元舆情智能检测系统设计
屈翼展伍建昌陈新
中国人民解放军空军预警学院
摘要：随着近年诸如网络自媒体、个人公众号的迅猛发展,其安全形势日益严峻。用户在浏览各种网络信息时,许多扰乱公众视线的信息也层出不穷。这些信息既可能包含恶意篡改图像内容,也可能包含负面文字信息,它们在某种程度上会影响公众的判断,造成不必要的社会恐慌。为此,本系统将深度学习技术以及经典算法检测用于负面评论和恶意篡改图像信息的检测,对负面的文本、图像信息做到及时发现、准确检测,以解决用户在浏览网页时无法判断信息内容是否含有虚假图像以及负面评论的问题。
关键词：互联网; 智能检测; 深度学习;
分类号：TP391.1;TP18

总结

从2019年至今中国知网共收录期刊180篇，其中英文较多，观察前40篇主要是，各种场景文本的识别，和文本识别算法的优化和利用文字检索找出书或其他一些文字信息中的关键字用来预防问题的发生类似于抑郁症。