Дисертації з теми "Features of video information"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся з топ-50 дисертацій для дослідження на тему "Features of video information".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Переглядайте дисертації для різних дисциплін та оформлюйте правильно вашу бібліографію.
Asghar, Muhammad Nabeel. "Feature based dynamic intra-video indexing." Thesis, University of Bedfordshire, 2014. http://hdl.handle.net/10547/338913.
Повний текст джерелаSjöblom, Mattias. "Investigating Gaze Attraction to Bottom-Up Visual Features for Visual Aids in Games." Thesis, Blekinge Tekniska Högskola, Institutionen för kreativa teknologier, 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:bth-12862.
Повний текст джерелаGurrapu, Chaitanya. "Human Action Recognition In Video Data For Surveillance Applications." Queensland University of Technology, 2004. http://eprints.qut.edu.au/15878/.
Повний текст джерелаJohansson, Henrik. "Video Flow Classification : Feature Based Classification Using the Tree-based Approach." Thesis, Karlstads universitet, 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:kau:diva-43012.
Повний текст джерелаHITS, 4707
Копинець, Валеріян Валеріянович. "Особливості оброблення відеоінформації для мультимедійних видань мистецького спрямування". Master's thesis, КПІ ім. Ігоря Сікорського, 2020. https://ela.kpi.ua/handle/123456789/39436.
Повний текст джерелаExplanatory note to the master's dissertation on "Features of video information processing for multimedia publications in the field of art", contains 72 pages, drawings 22, tables 24, literature sources 16. Integrated video tour in 360 format based on augmented reality technology. The main features of video information processing are investigated. An electronic and physical copy (page) of the publication was made for demonstration. Feasibility study of the project is given. The payback period of the project and the startup plan are calculated.
Šabatka, Pavel. "Vyhledávání informací." Master's thesis, Vysoké učení technické v Brně. Fakulta informačních technologií, 2010. http://www.nusl.cz/ntk/nusl-237213.
Повний текст джерелаGrinberg, Michael [Verfasser]. "Feature-Based Probabilistic Data Association for Video-Based Multi-Object Tracking / Michael Grinberg." Karlsruhe : KIT Scientific Publishing, 2018. http://www.ksp.kit.edu.
Повний текст джерелаMohanna, Farahnaz. "Content based video database retrieval using shape features." Thesis, University of Surrey, 2002. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.250764.
Повний текст джерелаNallabolu, Adithya Reddy. "Unsupervised Learning of Spatiotemporal Features by Video Completion." Thesis, Virginia Tech, 2017. http://hdl.handle.net/10919/79702.
Повний текст джерелаMaster of Science
Černý, Petr. "Vyhledávání ve videu." Master's thesis, Vysoké učení technické v Brně. Fakulta informačních technologií, 2012. http://www.nusl.cz/ntk/nusl-236590.
Повний текст джерелаJianshu, Chao [Verfasser], Eckehard [Akademischer Betreuer] Steinbach, and Matteo [Akademischer Betreuer] Cesana. "Feature-preserving image and video compression / Chao Jianshu. Betreuer: Eckehard Steinbach. Gutachter: Matteo Cesana ; Eckehard Steinbach." München : Universitätsbibliothek der TU München, 2016. http://d-nb.info/1088725104/34.
Повний текст джерелаChen, Bo. "Deep learning of invariant spatio-temporal features from video." Thesis, University of British Columbia, 2010. http://hdl.handle.net/2429/27651.
Повний текст джерелаComstedt, Erik. "Effect of additional compression features on h.264 surveillance video." Thesis, Mittuniversitetet, Avdelningen för informationssystem och -teknologi, 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:miun:diva-30901.
Повний текст джерелаYilmazturk, Mehmet Celaleddin. "Online And Semi-automatic Annotation Of Faces In Personal Videos." Master's thesis, METU, 2010. http://etd.lib.metu.edu.tr/upload/12611936/index.pdf.
Повний текст джерелаClark, Angus Alistair. "Region classification for the interpretation of video sequences." Thesis, University of Bristol, 2000. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.302167.
Повний текст джерелаMoroz, Viktor. "Features of information security in martial law." Thesis, National Aviation University, 2021. https://er.nau.edu.ua/handle/NAU/53731.
Повний текст джерелаThelin, Robert. "Interactive Video in Online Education : Evaluation of Current Video Delivery Systems for Interactive Features Defined in Literature." Thesis, Umeå universitet, Institutionen för datavetenskap, 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-128304.
Повний текст джерелаGupta, Ankur. "Using line and ellipse features for rectification of broadcast hockey video." Thesis, University of British Columbia, 2010. http://hdl.handle.net/2429/30490.
Повний текст джерелаGuan, Hao. "Local features, structure-from-motion and view synthesis in spherical video." Thesis, University of York, 2016. http://etheses.whiterose.ac.uk/17414/.
Повний текст джерелаIyengar, Giridharan Ranganathan 1969. "Information theoretic measures for encoding video." Thesis, Massachusetts Institute of Technology, 1995. http://hdl.handle.net/1721.1/61531.
Повний текст джерелаIsaieva, O. A., and О. Г. Аврунін. "Video dermoscopy study of the skin." Thesis, Liverpool, United Kingdom, 2019. http://openarchive.nure.ua/handle/document/10265.
Повний текст джерелаBanda, Nagamani. "Adaptive video segmentation." Morgantown, W. Va. : [West Virginia University Libraries], 2004. https://etd.wvu.edu/etd/controller.jsp?moduleName=documentdata&jsp%5FetdId=3520.
Повний текст джерелаTitle from document title page. Document formatted into pages; contains vi, 52 p. : ill. (some col.). Includes abstract. Includes bibliographical references (p. 50-52).
Kovesi, Peter. "Invariant measures of image features from phase information." University of Western Australia. Dept. of Computer Science, 1996. http://theses.library.uwa.edu.au/adt-WU2003.0006.
Повний текст джерелаKorpinen, K. P. (Kalle-Pekka). "Projektinhallinan video yliopisto-opetuksessa." Master's thesis, University of Oulu, 2014. http://urn.fi/URN:NBN:fi:oulu-201405241497.
Повний текст джерелаCastagno, Roberto. "Video segmentation based on multiple features for interactive and automatic multimedia applications /." Lausanne : Ecole polytechnique fédérale, 1998. http://library.epfl.ch/theses/?nr=1894.
Повний текст джерелаManiccam, Suchindran S. "Image-video compression, encryption and information hiding /." Online version via UMI:, 2001.
Знайти повний текст джерелаWampler, Christopher. "Information leakage in encrypted IP video traffic." Thesis, Georgia Institute of Technology, 2014. http://hdl.handle.net/1853/54287.
Повний текст джерелаMadurga, Martín-Serrano Juan Lucas. "A method for a small web site to add some sharing features." Thesis, Linköping University, Department of Science and Technology, 2008. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-11006.
Повний текст джерелаThe aim of this diploma work is to compare and evaluate different methods to enable small web sites to make available videos which will be shared the users. Storage and bandwidth problems of small web sites are taken into account. The requirements of the system were defined as: configurable, feasible, easy use and easy to integrate. Existing solutions and different implementation alternatives are analyzed. With a tool based upon a CMS, using recordings residing on powerful dedicated video providers and with AJAX “technology”, the criteria are fulfilled. As part of the investigation, a prototype tool based upon TYPO3 has been implemented.
Head, Milena M. "User interface features, facilitating information access and decision making." Thesis, National Library of Canada = Bibliothèque nationale du Canada, 1998. http://www.collectionscanada.ca/obj/s4/f2/dsk1/tape10/PQDD_0007/NQ42743.pdf.
Повний текст джерелаHead, Milena M. "User interface features : facilitating information access and decision making /." *McMaster only, 1997.
Знайти повний текст джерелаRudenko, Viktoria, Виктория Александровна Руденко, and Вікторія Олександравна Руденко. "Features of structure and regulation of the information market." Thesis, Видавництво СумДУ, 2007. http://essuir.sumdu.edu.ua/handle/123456789/8355.
Повний текст джерелаKarphammar, Millie, and Jennifer Brettschneider. "eCRM features and customer loyalty : A qualitative study within the video streaming industry." Thesis, Umeå universitet, Företagsekonomi, 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-185356.
Повний текст джерелаPrice, Mathew. "Using colour features to classify objects and people in a video surveillance network." Master's thesis, University of Cape Town, 2004. http://hdl.handle.net/11427/5121.
Повний текст джерелаVisual tracking of humans has proved to be an extremely challenging task for computer vision systems. One idea towards a development of these systems is the incorporation of colour. Often colour appearance of a person can provide enough information to identify an object or person in the short-term. However, the imprecise nature of colour measurements typically encountered in image processing has limited their use. This thesis presents a system which uses colour appearances of objects and people for tracking across multiple camera views in a digital video surveillance network.
Salam, Sazilah. "VidIO : a model for personalized video information management." Thesis, University of Southampton, 1996. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.242411.
Повний текст джерелаBarannik, Vlad, Y. Babenko, S. Shulgin, and M. Parkhomenko. "Video encoding to increase video availability in telecommunication systems." Thesis, Taras Shevchenko National University of Kyiv, 2020. https://openarchive.nure.ua/handle/document/16582.
Повний текст джерелаMeyn, Anselm Joseph. "Mobile Video Crawler : Implementing a video streaming Quality of Experience measurement system." Thesis, KTH, Skolan för informations- och kommunikationsteknik (ICT), 2014. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-177326.
Повний текст джерелаAnvändningen av mobila tjänster som utnyttjar video streaming växer i en snabb takt. I och med att antalet webbtjänster som utnyttjar video streaming ökar, finns det även ett växande intresse för att förstå Quality of Experience (QoE) för användningen av dessa tjänster, samt olika faktorer som inverkar på tjänstens QoE. QoE har en direkt inverkan på kundbevarande, vilket innebär att dålig QoE leder till förlorade intäkter för videotjänster. En videotjänsts kundupplevelse kan objektivt mätas med variabler som t.ex. antalet buffringspauser under tiden som videon spelas, samt tiden det tar för en video att börja spela. Att samla dessa data för analys på en större skala utgör en stor utmaning. I denna avhandling diskuterar vi arbetet som gjorts inom utvecklingen av ett system som samlar in och analyserar dessa variabler. Majoriteten av arbetet omfattar utvecklandet av en mobilapplikation som automatiskt spelar upp en lista av videon vid inställbara intervall, och under uppspelningen bandar in intressanta händelser så som statistik över systemet och nätverksanvändning, samt videospelarens händelser. Applikationen har blivit publicerad och kan användas från olika platser för att samla in en stor uppsättning data för detaljerad analys och modellering av QoE inom mobil Internet-video streaming.
Shah, Y. C. "Extraction of range information from stereo video images." Thesis, City University London, 1985. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.370923.
Повний текст джерелаAkinola, Mobolaji. "Intelligent side information generation in distributed video coding." Thesis, Open University, 2015. http://oro.open.ac.uk/42625/.
Повний текст джерелаRodrigues, Arturo Miguel Batista. "Coding of video with a single information plane." Master's thesis, Universidade de Aveiro, 2009. http://hdl.handle.net/10773/2103.
Повний текст джерелаAs actuais normas para codificação de vídeo, tais como os MPEG2/4 ou H.263/4, foram desenvolvidas para codificação de vídeo com cor. A informação de cor é representada usando um espaço apropriado, como, por exemplo, o YCbCr. Estes espaços de cor são constituídos por três planos: um para a dominância (no exemplo dado, o Y) e dois para a informação de crominância (neste caso, o Cb e o Cr). Contudo, há aplicações onde a informação a codificar é composta apenas por um plano de informação que pode, por exemplo, representar níveis de cinzento em imagem médica, ou índices para tabelas de cores. A motivação desta tese prende-se com dois factos: a produção de imagens médicas em formato digital estar a crescer, impondo técnicas eficazes para o tratamento e a compressão de dados e, embora os modelos de cor indexada sejam há muito utilizados para representar imagens, não têm sido convenientemente explorados em vídeo. Com esta dissertação pretende-se investigar novas estratégias de compressão sem perdas que explorem a redundância entre imagens consecutivas que caracterizam estas modalidades de imagem. Portanto, ao longo do trabalho implementou-se dois codificadores de vídeo para um só plano de informação, baseados num modelo híbrido. Um deles utiliza codificação de Golomb e o outro codificação aritmética, estudando-se assim a eficácia de cada um, quer para a escala de cinzentos, quer para vídeos com tabela de cores indexadas. Adicionalmente, para vídeos de cor indexada, implementou-se um algoritmo de reordenação da tabela de cores, o que torna a codificação mais eficaz. ABSTRACT: The current standards for video encoding, such as MPEG2/4 or H.263/4, have been developed for encoding video with color. The color information is represented using an appropriate space, such as YCbCr. These color spaces are made of three planes: one for luminance (in the given example, the Y) and two for the chrominance information (in this case, the Cb and Cr). However, there are applications where the information lies in a single information plane that may, for example, represent shades of gray (medical imaging) or indexes to color tables (color indexed video). The motivation of this thesis is related with two points: the production of medical images in digital format has been growing, imposing efficient techniques for the treatment and compression of data and, although color indexed models have been used for a long time to represent images, it has not been adequately explored in video. With this thesis, we intended to investigate new strategies for lossless compression which exploits the redundancy between consecutive images that characterize these types of images. Therefore, during this work, it has been implemented two video encoders with one information plane, based on a hybrid model. One of them uses Golomb codes and the other arithmetic coding. It has been studied the efficiency of each one, both using gray scale and color indexed videos. In addition, for color indexed videos, it has been implemented a palette reordering algorithm, making the encoding more efficient.
McCarthy, Timothy Mortimer Mark. "Integrating aerial video with G.I.S." Thesis, Birkbeck (University of London), 1999. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.324964.
Повний текст джерелаHu, Wei Shu. "Mining product features from online reviews." Thesis, University of Macau, 2010. http://umaclib3.umac.mo/record=b2148259.
Повний текст джерелаVan, der Haar Dustin Terence. "Face recognition-based authentication and monitoring in video telecommunication systems." Thesis, 2012. http://hdl.handle.net/10210/5024.
Повний текст джерелаA video conference is an interactive meeting between two or more locations, facilitated by simultaneous two-way video and audio transmissions. People in a video conference, also known as participants, join these video conferences for business and recreational purposes. In a typical video conference, we should properly identify and authenticate every participant in the video conference, if information discussed during the video conference is confidential. This prevents unauthorized and unwanted people from being part of the conference and exposing any confidential information during the video conference. Present existing video conferencing systems however, have problems in this area, resulting in some risks. These risks relate precisely to the lack of facilities to properly identify and authenticate participants, making it possible for unwanted/unauthorised participants to join the conference or masquerade as another participant. It is especially a problem, when facilitators or organisers are the only participants that know the authorised participants, or participants allowed in a video conference. In this dissertation, we review the risks that are present in video conferencing, and create a security system, (called BioVid) that mitigates the identification and authentication risks in video conferences. BioVid uses a Speeded-Up Robust Features or SURF-based face recognition approach, to identify and authenticate any participant in a video conference. BioVid continuously monitors the participants to check if masquerading has occurred and when it does detect an unauthorised participant, it informs the Service Provider. The Service Provider can then deal with the problem by either kicking the participant or asking the other participants to vote the unauthorised participant out of the video conference.
Ramalho, Pedro José Ascensão. "Feature Extraction and Object Classification in Video Sequences for Military Surveillance." Dissertação, 2019. https://hdl.handle.net/10216/122404.
Повний текст джерелаObject detection and recognition requires a learning system that can automatically identify a group of objects independently of the input data. To be able to perform this kind of identification, this system needs to previously analyze a large group of data, so it can memorize special features of different objects. This procedure it's called training and it's the first step in all the detection and recognition processes of machine learning. Although there are already many models that perform detection and recognition for a large group of objects, one of the goals of this project is to specify this identification into a small and special group of objects. This will be achieved by using transfer learning, that is a process that uses the knowledge gained by one of these models while solving one problem and applies it to a different one. Basically, it takes advantage of the feature extraction procedure outputs and use them to learn how to identify other kind of objects. Feature extraction is a group of processes with the goal of simplifying big groups of data by creating small sets of non-redundant information. These small groups are more manageable and can fully describe the original data set and, by using them, the resources necessary to analyse a large set of input data are decreased. In this context, the data to be analyzed will be captured by a camera implemented at a stationary point or in a vehicle. When dealing with the capture of visual information, it's normal that a large number of data is generated. So, it's important to analyze it efficiently and achieve relevant information identification. This dissertation focuses in military uses, therefore these operations are going to be used to automatically identify objects in the military field, that is, tanks, guns, people and vehicles (cars and trucks), achieving territorial surveillance.
Ramalho, Pedro José Ascensão. "Feature Extraction and Object Classification in Video Sequences for Military Surveillance." Master's thesis, 2019. https://hdl.handle.net/10216/122404.
Повний текст джерелаObject detection and recognition requires a learning system that can automatically identify a group of objects independently of the input data. To be able to perform this kind of identification, this system needs to previously analyze a large group of data, so it can memorize special features of different objects. This procedure it's called training and it's the first step in all the detection and recognition processes of machine learning. Although there are already many models that perform detection and recognition for a large group of objects, one of the goals of this project is to specify this identification into a small and special group of objects. This will be achieved by using transfer learning, that is a process that uses the knowledge gained by one of these models while solving one problem and applies it to a different one. Basically, it takes advantage of the feature extraction procedure outputs and use them to learn how to identify other kind of objects. Feature extraction is a group of processes with the goal of simplifying big groups of data by creating small sets of non-redundant information. These small groups are more manageable and can fully describe the original data set and, by using them, the resources necessary to analyse a large set of input data are decreased. In this context, the data to be analyzed will be captured by a camera implemented at a stationary point or in a vehicle. When dealing with the capture of visual information, it's normal that a large number of data is generated. So, it's important to analyze it efficiently and achieve relevant information identification. This dissertation focuses in military uses, therefore these operations are going to be used to automatically identify objects in the military field, that is, tanks, guns, people and vehicles (cars and trucks), achieving territorial surveillance.
Barbosa, Patrícia Margarida Silva de Castro Neves. "Human features detection in video surveillance." Master's thesis, 2016. http://hdl.handle.net/1822/46590.
Повний текст джерелаHuman activity recognition algorithms have been studied actively from decades using a sequence of 2D and 3D images from a video surveillance. This new surveillance solutions and the areas of image processing and analysis have been receiving special attention and interest from the scientific community. Thus, it became possible to witness the appearance of new video compression techniques, the transmission of audio and video in real-time, targeting identification and tracking objects in with complex environments. Traffic monitoring, automotive safety, people counting and activity recognition applications are examples. With the development of sensors, new opportunities arose to expand and advance this field. This dissertation presents an activity recognition system to recognize human motion. The system does not need optical markers or motion sensors. This human activity recognition system is divided in three stages: human segmentation, in an outside and inside environment; extraction of the human features; and classification models to detect the human actions. Therefore, the main objective in this work is to develop an algorithm to extract human features. This algorithm aims to develop a new representation and extraction method using a sequence of features in a skeleton silhouette. Mainly, the segmentation of humans is based on a previous work, centered on background subtraction. An algorithm is applied to convert the object captured in the video surveillance to a binary image using a skeleton algorithm. Afterwards, and based on the physical parameters of the human motion, it becomes possible to discover the principal features of the human skeleton, called physical features, head, hands and feet. The viability of using features detection in a human recognition system was tested and compared with other existing systems. The results point out that the system has good performance (8.96% of perfect match and the average rate was 68.65%). Nevertheless, in images where the features of the human body are covered, with umbrella or heavy coats for example, the system presents certain limitations. This process has a high execution speed and a low cost computational processing: average of 5910 µs with a standard deviation of 5650 µs. In the near future, classification models to detect the human actions will be included.
Algoritmos de reconhecimento de atividade humana foram estudados ativamente durante décadas, usando sequências de imagens em 2D e 3D de vídeo vigilância. Estas novas soluções de vídeo vigilância e as áreas de processamento e análise de imagens têm recebido especial atenção e interesse por parte da comunidade científica. Assim, tornou-se possível testemunhar a aparência de novas técnicas de compressão de vídeo, a transmissão de áudio e vídeo em tempo real, identificação de segmentação e rastreamento de objetos em ambientes complexos. Monitoramento de tráfego, segurança automóvel, contagem de pessoas e aplicações de reconhecimento de atividade são exemplos. Com o desenvolvimento de sensores, novas oportunidades surgiram para expandir e avançar neste campo. Esta dissertação apresenta um sistema de reconhecimento de atividade para reconhecer o movimento humano. O sistema não precisa de marcadores óticos ou sensores de movimento. Este sistema de reconhecimento de atividade humana divide-se em três fases: segmentação humana, num ambiente exterior e interior; Extração das características humanas; E modelos de classificação para detetar as ações humanas. Portanto, o objetivo principal deste trabalho trata-se de desenvolver um algoritmo para extrair características humanas. Este algoritmo tem como objetivo desenvolver uma nova representação e método de extração de características humanas, através do uso de uma silhueta em forma de esqueleto. A segmentação de seres humanos é baseada num trabalho anterior, centrado na subtração do plano de fundo. Um algoritmo é aplicado para converter o objeto capturado na vídeo vigilância, para uma imagem binária usando um algoritmo em forma de esqueleto. Posteriormente, e com base nos parâmetros físicos do movimento humano, torna-se possível descobrir as principais características do esqueleto humano, denominadas características físicas, cabeça, mãos e pés. A viabilidade do uso de deteção de características em um sistema de reconhecimento humano foi testada e comparada com outros sistemas. Os resultados indicam que o sistema tem bom desempenho (8.96% de correspondência exata e 68.65% de correspondência intermédia). No entanto, em imagens onde as características do corpo humano são cobertas, com guarda-chuva ou casacos pesados, por exemplo, o sistema apresenta certas limitações. Este processo tem uma alta velocidade de execução e um processamento computacional de baixo custo: média de 5910 μs com desvio padrão de 5650 μs. Num futuro próximo, serão incluídos modelos de classificação para detetar as ações humanas.
Tsou, Chih-Wei, and 鄒志偉. "Video Forgery Detection Using Combined Features." Thesis, 2009. http://ndltd.ncl.edu.tw/handle/45728420382667617866.
Повний текст джерела國立中正大學
資訊工程所
97
We propose a new approach for locating forged regions in a video sequence without prior information. We also propose an approach which can detect different types of forgeries. In our test sequences, we use some forgeries such as video inpainting, splicing, resample, and dynamic texture synthesis. We utilize residual quantization error to detect forgeries and join color filter array, the correlation of noise residual to enhance our approach. These three features are good at some types of forgeries. Because these features have complementariness, we choose these features to detect different types of forgeries. Finally, we use decision tree to combine these features.
Lin, Yung-Chieh, and 林永傑. "Video Object Segmentation Using Flow-Thread Features." Thesis, 2000. http://ndltd.ncl.edu.tw/handle/39984296827681619309.
Повний текст джерела國立臺灣大學
資訊工程學研究所
88
The object-oriented functionality becomes one of the most important issues for the video processing system nowadays. Especially in the new generation of the video coding standards, MPEG-4, a video object in a scene is an entity that a user is allowed to access and manipulate. The instance of a video object at a given time can be given by a sub-image in a video frame with an arbitrary contour. The goal of the video object segmentation is to extract the instances of the video objects in an image sequence. In general, methods for video object segmentation assume that the individual video object has consistent spatiotemporal information. To utilize both the spatial and temporal information, the proposed method in this thesis combines the results of single image segmentation with the flow-thread features. Here, a flow-thread is a series of pixels obtained by tracing the motion vectors along the image sequence. This thesis is organized into three parts: (1) the segmentation of single images, (2) the flow-thread construction, and (3) the flow-thread analysis and classification. In the segmentation of single images, we will introduce a popular morphological method, and compare its two algorithms using different approaches. In the flow-thread construction, a fast algorithm is used to refine the pixel-level motion vectors to the subpixel precision. This algorithm is based on the block matching and bilinear interpolation. After the motion vectors are refined, the flow-threads can be constructed by tracing the motion vectors. In the flow-thread analysis, the discrete Fourier transform and discrete wavelet transform are used for extracting the features of flow-threads. Finally, we use the conventional pattern recognition techniques to segment the scene into VOs based on the flow-thread features and regions obtained in the single image segmentation. Our method was used to segment some MPEG-4 test sequence, and the segmentation results are promising.
Lin, Chih-long, and 林志隆. "Content-based Video Retrieval with Multi Features." Thesis, 2014. http://ndltd.ncl.edu.tw/handle/ed84t6.
Повний текст джерела國立臺灣科技大學
電機工程系
102
With the advance of multimedia codec technology and communications, multimedia communications become one of the major information media with the aides of internet prevalence. Under this circumstance, image and video data over the Internet contribute to the sea of media and how to search user desired media contents from the sea of media becomes important. Content-Based Video Retrieval (CBVR) methods have been proposed to search user interested video clips, precisely and quickly. Among these researches, extracting image features for similarity measurement is widely adopted. However, adopting only one kind of feature to describe video contents cannot provide satisfactory retrieval results. In general, more than one kind of image/video features are extracted to for efficient video retrieval. How to efficiently integrate different kinds of image/video features is critical and challenging in improving the video retrieval performance. In this thesis, we proposed to integrate color, texture and SIFT-BOW (Bag of Word) image features to describe one video clip. These three features not only can describe the global image feature, but also local region ones. In our experiments, the color histogram difference is used to measure similarity for video scene cut. These video scene cuts, video clips, are used as the basic media unit for description and retrieval. The average of image features within one media unit is used as the representative feature for the video clip. To perform retrieval, the feature of one query image/video is extracted and its similarity to each representative feature of one video clip in a database is calculated to perform similarity ranking. For comparisons, the video retrieval performance that adopts only one feature is implemented. In addition, the one proposed by Y. Deng [10] that adopts more than one feature for video retrieval is also carried out for comparisons. Experiments showed that the proposed CBVR method outperforms the previous method by 38.7% in the PR rate. Performing CBVR by multiple features also improves on the PR performance as compared to retrieval by single feature.
Wu, Jui-Chen, and 吳瑞珍. "Salient Features Extraction for Video Content Analysis." Thesis, 2009. http://ndltd.ncl.edu.tw/handle/30681624290573307145.
Повний текст джерела元智大學
電機工程學系
97
Multimedia can be widely used due to the progress of science and technology. An automatic object detection and analysis system is necessary that we can ability transmits, store and retrieve for video data. To achieve these goals, there have been many approaches proposed for detecting object for a plan or visual median. Hence, having a detailed description of video content analysis can provide rich information for user. In this thesis, we will propose two novel salient features extraction schemes for text detection and recognition in images or video sequences, vehicle orientation analysis and vehicle retrieval from image databases. First, the morphology-based scheme can be used to find out high contrast region with their background. The method is invariant under different lighting, scaling, and viewing conditions. As a result of text often having high contrast with their background, all possible candidate regions will be extracted. Finally, the geometric properties can be used to detect text line from images. After extracting, we will recognize license plate from video sequences. Moreover, to fast and effectively analyze vehicles from image databases, we proposed the “eigen” color extraction scheme to detect possible vehicle regions from cluttered images. The model can efficiently separate foreground pixels from the cluttered images even under different lighting conditions. After extracting candidates regions, we will define some descriptors to achieve vehicle orientation analysis and vehicle retrieval system. Experimental results reveal the superior performances in text extraction, license plate recognition, vehicle orientation analysis, and vehicle retrieval.
Chen, Wei-Chung, and 陳韋均. "Rushes Video Summarization by Audio-filtering Visual Features." Thesis, 2012. http://ndltd.ncl.edu.tw/handle/43004645505593306250.
Повний текст джерела國立雲林科技大學
資訊工程系碩士班
100
In this paper, we propose a video summarization system for analyzing basketball game videos. In contrast to previous video analysis technologies employing only visual and motion features to do video filtering, we add audio features to do video summarization in the system. First, we extract replay highlights by special effect detection. Next, we filter landscape shots using color range pixel and fast motion activity. Then, the corresponding audio features extracted from these landscape shots are used to identify landscape shot highlights by an SVM. Finally, we integrate the replay highlights and landscape shot highlights to complete the video summarization. From the experimental results, we find that the accuracy on the special effect detection, landscape shot extraction, and landscape shot highlight detection is very high. Thus, the final video summarization has high recall values on highlight extraction.