P. Salembier and F. Marques, "Region-based representations of image and video: segmentation tools for multimedia services," IEEE Trans. CASVT 9#8#, 1147--1169 #1999#.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....contours, and special features like corners, edges, vertices, etc. There are approaches that try to distinguish shadows from the objects. There are also methods that are proposed for the compressed domain rather than pixel domain. One method for extraction of objects extracts region information [80, 33, 34, 26, 100, 106, 32, 60, 101, 35]. A generic partition tree for the regions that appear in the image is generated in [80] This generic partition tree can be used for unsupervised and supervised spatial segmentation, region based coding, semantic region segmentation, and motion spatial segmentation. The binary partitioning tree ....

....the objects. There are also methods that are proposed for the compressed domain rather than pixel domain. One method for extraction of objects extracts region information [80, 33, 34, 26, 100, 106, 32, 60, 101, 35] A generic partition tree for the regions that appear in the image is generated in [80]. This generic partition tree can be used for unsupervised and supervised spatial segmentation, region based coding, semantic region segmentation, and motion spatial segmentation. The binary partitioning tree based on regions are also used in [33] and they use it for retrieval from databases. A ....

P. Salembier and F. Marques. Region-based representations of image and video: Segmentation tools for multimedia services. IEEE Transactions on Circuits and Systems for Video Technology, 9(8):1147--1167, December 1999.

....of these methods. Moving objects could also be used for content description in MPEG 7 applications. Various approaches have been proposed for video or spatio temporal segmentation. An overview of segmentation tools, as well as of region based representations of image and video, are presented in [6]. The video object extraction could be based on change detection and moving object localisation, or on motion field segmentation, particularly when the camera is moving. Our approach is based exclusively on change detection. The costly and potentially inaccurate motion estimation process is not ....

P.Salembier and F. Marques, "Region-based Representations of Image and Video: Segmentation Tools for Multimedia Services," IEEE Trans. on Circuits and Systems for Video Technology. Vol. 9, pp. 1147--1169, Dec. 1999.

....of these methods. Moving objects could also be used for content description in MPEG 7 applications. Various approaches have been proposed for video or spatio temporal segmentation. An overview of segmentation tools, as well as of region based representations of image and video, are presented in [17]. The video object extraction could be based on change detection and moving object localisation, or on motion field segmentation, particularly when the camera is moving. Our approachisbased exclusively on change detection. The costly and potentially inaccurate motion estimation process is not ....

P. Salembier and F. Marques. Region-based representations of image and video: Segmentation tools for multimedia services. IEEE Trans. on Circuits and Systems for VideoTechnology, 9:1147--1169, Dec. 1999.

....image retrieval efficiency [2] The additional feature information regarding neighbor relationships between objects can be used to further improve image retrieval for image queries where spatial information about objects in the image is key to effective image recall. For MPEG 4 video compression [3], and other content based video compression schemes, image segmentation is used to divide scenes up into objects. By performing motion estimation and compensation on objects, the minimal temporal variations within each object enables higher compression ratios. The added benefit of spatial ....

P. Salembier and F. Marques, "Region-based representations of image and video: Segmentation tools for multimedia services, " IEEE Trans. Circuits Syst. Video Technol., vol. 9, pp. 1147--1169, December 1999.

....image retrieval efficiency [98] The additional feature information regarding neighbor relationships between objects can be used to further improve image retrieval for image queries where spatial information about objects in the image is key to effective image recall. For MPEG 4 video compression [60], and other content based video compres sion schemes, image segmentation is used to divide scenes up into objects. By per forming motion estimation and compensation on objects, the minimal temporal vari ations within each object enables higher compression ratios. The added benefit of spatial ....

P. Salembier and F. Marques. Region-based representations of image and video: Segmentation tools for multimedia services. IEEE Trans. Circuits Syst. Video Technol., 9:1147-1169, December 1999. 101

....object. Working on these assumptions we develop a spatial segmentation algorithm which combines colour and motion information to identify the coherent moving objects in the shot. The algorithm begins with a colour segmentation of a single frame, generated using the spanning tree technique of [7]. This produces a binary tree representation of the frame; each node represents a connected region with the root representing the whole frame. The children of each node represent a split along the colour boundaries in that region. 2.2.1 Motion estimation Motion parameters are computed for a ....

P. Salembier and F. Marques. Region-based representations of image and video: segmentation tools for multimedia services. IEEE Transactions on Circuits and Systems for Video Technology, 9(8):1147--1167, December 1999.

....Object based encoding for natural videos, however, is not easy to achieve. Objects need to be segmented, and tracked so that each object can be encoded independently. Content based image and video retrieval has emerged as a significant field of computer vision and multimedia in recent years [10, 49, 48]. It has shown so much promise that MPEG is developing MPEG 7 to standardize ways to describe various types of audiovisual information including video data [44, 40] The abundance of images and videos on the Internet and the World Wide Web (WWW) makes it very di#cult for users to find desired ....

....as a topographic surface. It detects the minima of the gradient of the gray level image and grows these minima according to the gradient values much like a flooding process. Points of contact between the propagation originating from di#erent minima are defined as the boundaries of the regions [48]. Many other approaches have also been proposed. Statistical methods have been very popular. In fact, Blobworld is very much a statistical method which models data as a mixtures of Gaussians. Other popular models used include Fourier statistics, Covariance statistics, label statistics and ....

[Article contains additional citation context not shown here]

P. Salembier and F. Marques. Region-based representations of image and video: segmentation tools for multimedia services. IEEE Trans. on Circuits and Systems for Video Technology, 9(8):1147--1169, 1999.

.... on global similarities, such as color, texture, and motion (e.g. 8] 12] 16] 17] 22] 23] More recently, a separate set of works has started to address localized, regional representations that enable spatio temporal segmentation for object based video retrieval (e.g. 9] 5] 4] [19]) Spatio temporal segmentation has been a very challenging research problem, and many algorithms are proposed in the literature [10] 5] 16] 20] Many approaches use optical flow methods [14] to estimate motion vectors at the pixel level, and then cluster pixels into regions of coherent 2 ....

P. Salembier and F. Marques. Region-based representations of image and video: Segmentation tools for multimedia services. IEEE Trans. on Circuits and Systems for Video Technology, 9(8):1147--1168, 1999.

....of COST 211 quat is also gratefully acknowledged. # For general image collections, there are currently no systems that automatically classify images or recognize the objects they contain. In order to overcome these problems, a region based color indexing and retrieval algorithm is presented [6, 7]. As a basis for the indexing, a novel K Means segmentation algorithm is used, modified so as to take into account the coherence of the regions. A new color distance is also defined for this algorithm. The # # # # # # colour space is used [8] which is related to the CIE 1931 XYZ standard ....

P. Salembier and F. Marques, "Region-Based Representations of Image and Video: Segmentation Tools for Multimedia Services," IEEE Trans. Circuits and Systems for Video Technology, vol. 9, no. 8, pp. 1147--1169, December 1999.

....database in real time. In order to achieve reasonable search speed, some form of object segmentation is required so that object features can be retrieved during preprocessing (off line) and stored in the database for matching. It is likely that a perfect image segmentation is impossible to attain [13, 14]. Content based retrieval can be achieved with a more relaxed form of image segmentation [14] with fuzzy similarity measure. The Blobworld system [15] allows object query by the user selecting an image region (Query by Example) Image segmentation uses polarity, anisotropy and normalized texture ....

....is required so that object features can be retrieved during preprocessing (off line) and stored in the database for matching. It is likely that a perfect image segmentation is impossible to attain [13, 14] Content based retrieval can be achieved with a more relaxed form of image segmentation [14], with fuzzy similarity measure. The Blobworld system [15] allows object query by the user selecting an image region (Query by Example) Image segmentation uses polarity, anisotropy and normalized texture contrast for each pixel, and color in L a b space. Pixels are grouped assuming a mixture of ....

P. Salembier and F. Marques. Region-based representations of image and video: segmentation tools for multimedia services. IEEE Trans. on Circuits and Systems for Video Technology, 9(8):1147--1169, 1999.

.... on global similarities, such as color, texture, and motion (e.g. 6] 10] 12] 13] 17] 18] More recently, a separate set of works has started to address localized, regional representations that enable spatio temporal segmentation for objectbased video retrieval (e.g. 7] 3] 2] [14]) Spatio temporal segmentation has been a very challeng1 ing research problem, and many algorithms are proposed in the literature [8] 3] 12] 15] Many approaches use optical flow methods [11] to estimate motion vectors at the pixel level, and then cluster pixels into regions of coherent ....

P. Salembier and F. Marques. Region-based representations of image and video: Segmentation tools for multimedia services. IEEE Trans. on Circuits and Systems for Video Technology, 9(8):1147--1168, 1999.

....In our scheme this is done by reducing the values of a l 1 #l 2 [i# j] for i j : even or i j : odd, depending on which sum is greater. 4 Segmentation algrithm Segmentation methods for 2D images may be divided primarily into region based and boundary based methods. Region based approaches [5] rely on the homogeneity of spatially localised features suchasgraylevel intensity and texture. Region growing and split and merge techniques also belong to the same category. On the other hand, boundary based methods use primarily gradient information to locate object boundaries. In this paper, a ....

P. Salembier and F. Marques, "Region-Based Representations of Image and Video: Segmentation Tools for Multimedia Services," IEEE Trans. Circuits and Systems for Video Technology, vol. 9, no. 8, pp. 1147--1169, December 1999.

....[1, 2] Arguably, one of the most useful searching methods is by object model. Global image features do not lend themselves well to object retrieval. Background objects can heavily affect the global feature vectors. Recently, attempts have been made to use image segmentation for object indexing [3] with some success. There is no system however that retrieves objects based on visual contents or models with high speed and accuracy. In our C BIRD (Content Based Image Retrieval from Digital libraries) system [4] we use a technique for coarse localization of image features into locales. Locales ....

P. Salembier and F. Marques. Region-based representations of image and video: segmentation tools for multimedia services. IEEE Trans. on Circuits and Systems for Video Technology,9(8):1147--1169,1999.

....etc. However, one of the most useful searching methods is by object model. Global image features do not lend themselves well to object retrieval, since background objects can greatly affect the global feature vectors. While image segmentation for object representation has been widely studied [4], the problem is arguably ill defined and a good segmentation is often impossible to achieve. In CBIRD (Content Based Image Retrieval from Digital libraries) 5] we developed a technique for coarse localization of image features into locales. The set of image locales may be overlapped and or ....

P. Salembier and F. Marques. Region-based representations of image and video: segmentation tools for multimedia services. IEEE Trans. on Circuits and Systems for Video Technology,9(8):1147--1169,1999.

....the major techniques for video and image indexing can be reviewed in [1] For video indexing applications, the initial phase generally consists in structuring the original content. A classical structuring approach consists in detecting shots in a video sequence and to group them into scenes (see [2, 3, 4] and the references herein) The description of shots is often based on key frames. For instance in [5] several key frames are used to represent a set of shots and to browse them. These key frames are also indexed using standard techniques for still images. Other methods for shot representation ....

P. Salembier and F. Marques, "Region-based representations of image and video: segmentation tools for multimedia services, " IEEE Trans. on Circuits and Systems for Video Technology, vol. 9, no. 8, pp. 1147--1169, December 1999.

No context found.

P. Salembier and F. Marques, "Region-based representations of image and video: segmentation tools for multimedia services," IEEE Trans. CASVT 9#8#, 1147--1169 #1999#.

No context found.

P. Salembier and F. Marques, "Region-based representations of image and video: Segmentation tools for multimedia services," IEEE Trans. Circuits Syst. Video Technol., vol. 9, no. 8, pp. 1147--1169, 1999.

No context found.

P. Salembier and F. Marqus, "Region-based representations of image and video: Segmentation tools for multimedia services," IEEE Trans. Circuits Syst. Video Technol., vol. 9, pp. 1147--1169, Dec. 1999.

No context found.

P. Salembier and F. Marques, "Region-based representations of image and video: Segmentation tools for multimedia services," IEEE Trans. Circuits Syst. Video Technol., vol. 9, pp. 1147--1169, Dec. 1999.

No context found.

Salembier, P. and F. Marques, Region-based Representations of Image and Video: Segmentation Tools for Multimedia Services, IEEE Transactions on Circuits and Systems for Video Technology, vol. 9, no. 8, pp. 11471169, 1999.

No context found.

P. Salembier and F. Marques, "Region-based representations of image and video: Segmentation tools for multimedia services," IEEE Trans. Circuits Syst. Video Technol., vol. 9, pp. 1147--1169, Dec. 1999.

No context found.

P. Salembier and F. Marques, "Region-Based Representations of Image and Video: Segmentation Tools for Multimedia Services," IEEE Trans. Circuits and Systems for Video Technology, vol. 9, no. 8, pp. 1147--1169, December 1999.

No context found.

P. Salembier, F. Marques, "Region-based representations of image and video : Segmentation tools for multimedia services," IEEE Trans. Circ. Sys. Vid. Tech., Vol. 9(8), 1999, pp. 11471167.

No context found.

P. Salembier and F. Marques, "Region-Based Representations of Image and Video: Segmentation Tools for Multimedia Services," IEEE Trans. Circuits and Systems for Video Technology, vol. 9, no. 8, pp. 1147--1169, December 1999.

No context found.

P. Salembier and F. Marques. "Region-based representations of image and video: segmentation tools for multimedia services". IEEE Trans. Circ. Sys. Vid. Tech., 9(8), pp. 1147-1169, 1999.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC