Picard, R.W.: "Light-years from Lena: Video and Image Libraries of the Future", In the Proceedings, IEEE International Conference on Image Processing 1995, Vol.1, pp. 310-313  Home/Search   Document Details and Download   Summary   Related Articles   Check

....Each such unit contains several temporally consecutive events, which are logically interrelated and belong to the same larger part of the story. 1 Introduction In the fast growing area of video databases, the problem to efficiently retrieve stored video material remains one of the key issues [8]. The efficiency of the retrieval procedure determines, at the same time, its user friendliness. To reach a high efficiency level, special attention has to be given to development of suitable methods for abstracting large amounts of video material and presenting it to the user in a logical, ....

Picard, R.W.: "Light-years from Lena: Video and Image Libraries of the Future", In the Proceedings, IEEE International Conference on Image Processing 1995, Vol.1, pp. 310-313

....is the management of large amounts of information. Though solutions exist for text oriented databases (e.g. SQL) methods for browsing, querying and organization of visual information, such as images, graphics and video, and their linking to textual information are still in their infancy [12]. The MPEG 7 standardization platform [6] addresses ways of representing visual information using a standard set of descriptors, such that it can be used effectively in professional and consumer applications and especially in user interfaces. The question as how to automatically obtain these ....

Picard R.W.: "Light-years from Lena: Video and Image Libraries of the Future", Proceedings of IEEE ICIP 1995, Vol.1, pp. 310-313.

....is either performed using histogram comparison [6] or thanks to a graph similarity measure [9] Promising results are presented along with possible work extention. 1. INTRODUCTION The topic of recognition of objects from large libraries of images has attracted much interest over the past decade [1, 2, 3, 4]. Most of the literature has focussed on using lowlevel image characteristics such as color [1] texture [5] or local feature orientation [3] for the purposes of recognition. One of the most efficient ways to realize recognition is to encode the distribution of image characteristics in a histogram ....

R. W. Picard, "Light-years from Lena: Video and image libraries of the future," ICIP'95, vol. 1, pp. 310--313, 1995.

....video content must be developed. The heterogeneity of the content of video in the general sense leads to think that there is no one general model, so that different models, which could cooperate in a general purpose framework, should be developed for different kinds of contents [4] Picard [5] suggested the possible key challenges within this research topic: finding good models for describing visual content in the sense of both compression and content access and finding good measures of visual similarity. A model for describing video content, while giving support for both perceptual ....

R. W. Picard (1995) Light-years from Lena: Video and image libraries of the future. In: Proceedings of the International Conference on Image Processing, Washington DC, October, Vol. I, pp. 310--313.

....must be developed. The heterogeneity of the content of video in the general sense leads to think that there is no one general model, so that different models, which could cooperate in a general purpose framework, should be developed for different kinds of contents [11] Picard suggested in [12] what possibly the key challenges within this research topic are: finding good models for describing visual content in the sense of both compression and content access and finding good measures of visual similarity. A model for describing video content, while giving support for both perceptual and ....

R. W. Picard. Light-years from Lena: Video and image libraries of the future. In Proc. Int. Conf. on Image Processing, Washington DC, October 1995.

....the best match. In typical applications a number of top matches with rankordered similarities to the query pattern will be retrieved. Comparison in the feature space should preserve visual similarities between patterns. This is an important but difficult problem in contentbased image retrieval [17]. Towards this objective, a neural network algorithm to learn the pattern similarity in the feature space is proposed (Section 3) This approach uses training data containing the pattern similarity information (provided by human indexers) to partition the feature space into 3 many visually ....

R. W. Picard, "Light-years from lena: video and image libraries of the future," Proceedings of IEEE Int. Conf. on Image Processing, vol. I, pp. 310-313, Oct. 1995.

....that while the histogrambased method is most sensitive to the addition of line segments and end point position errors, the set based method is most sensitive to line dropout. 1 Introduction The recognition of objects from large libraries is a problem of pivotal importance in image retrieval [9, 7, 6, 1]. The topic has attracted massive interest over the past decade. Most of the literature has focussed on using low level image characteristics such as colour [9] texture [2] or local feature orientation [6] for the purposes of recognition. One of the most efficient ways to realise recognition is ....

R. W. Picard. Light-years from lena: Video and image libraries of the future. IEEE ICIP'95, 1:310--313, 1995.

....index of line structure# it is robust both to image distortion an to the variable quality of input line segmentation. 1 Introduction Pivotal to the effectiveness of multi media or hypertext systems is the ability to accurately retrieve images from a large database on the basis of contents [1, 2, 3, 4, 5]. The crucial ingredient of the underlying image database management system is an accurate way of indexing according not only to the frequency of salient image features but also to meaningful relational structure between the features. Ideally,such content based indices should be capable of ....

....histogram distance measure. Finally, section 6 presents some conclusions and suggests directions for future investigation. 2 The Application Domain The problem of accurately and rapidly performing content based retrieval of images is very complex and is currently an active area of research[1, 2, 3, 4, 5]. The pictorial information used in these studies have invariably represented quite diverse scenes in which the different images are relatively salient with respect to one another. In this paper we consider the problem of indexing into a database of images having relatively uniform appearance. The ....

R. W. Picard, "Light-years from lena: Video and image libraries of the future," Int. Conf. on Image Proc., vol. 1, pp. 310--313, 1995.

....control by using a pen as the tool for drawing object outlines 5. Allow for incremental annotation of image material by storing user outline queries and annotations 6. Start with a limited content domain to evaluate these concepts 2. 1 Focus on object based representations and queries Picard[5] makes a distinction between (a) subject, object and action search, b) syntactic search (layout of images, breakpoints in video streams) c) mood related search, and the category (d) I know what I m looking for when I see it . Although looking for particular objects in images ( all horses , ....

Picard, R.W.: Light-years from Lena: Video and Image Libraries of the Future Proceedings of the International Conference on Image Processing (ICIP), Oct '95, Washington DC, USA. Vol I (1995) 310--313

....both the effectiveness of our representation and the advantage that it offers in overcoming local shape ambiguities. 1. INTRODUCTION Pivotal to the effectiveness of multi media or hypertext systems is the ability to accurately retrieve images from a large database on the basis of contents [1, 2, 3, 4, 5]. The crucial ingredient of the underlying image database management system is an accurate way of indexing according not only to the frequency of salient image features but also to meaningful relational structure between the features. Ideally, such content based indices should be capable of ....

R. W. Picard. Light-years from lena: Video and image libraries of the future. Int. Conf. on Image Proc., 1:310--313, 1995.

....index of line structure; it is robust both to image distortion an to the variable quality of input linesegmentation. 1 Introduction Pivotal to the effectiveness of multi media or hyper text systems is the ability to accurately retrieve images from a large database on the basis of contents [6, 7, 4, 10, 8]. The crucial ingredient of the underlying image database management system is an accurate way of indexing according not only to salient image features but also to meaningful relational structure between the features. Ideally, such content based indices should be capable of unambiguously ....

....histogram distance measure. Finally, section 6 presents some conclusions and suggests directions for future investigation. 2 The Application Domain The problem of accurately and rapidly performing content based retrieval of images is very complex and is currently an active area of research [6, 7, 4, 10, 8]. The pictorial information used in these studies have invariably represented quite diverse scenes in which the different images are relatively salient with respect to one another. In this paper we consider the problem of indexing into a database of images having relatively uniform appearance. The ....

R. W. Picard. Light-years from lena: Video and image libraries of the future. Int. Conf. on Image Proc., 1:310--313, 1995.

....besides the human consumer from congestion, new efficient access and compression structures have to be developed. Expecting the today s non interactive techniques soon to fail this challenge, a new generation of Visual Information Management Systems (VIMS) has been sketched by several authors[5, 6, 11]. VIMS are intended to seamlessly integrate visual, auditory, and symbolic information in many application domains, such as Geographic Information Systems, Engineering Visualization Systems, Medical Information Systems or Education Systems. The key paradigm of VIMS is content based image ....

R. W. Picard. Light--years from Lena: Video and image libraries of the future. Tech. report, M.I.T. Media Labaratory Perceptual Computing Section, 1995.

....segments pairs. Moreover, a sensitivity analysis reveals that segmentation errors do not affect the retrieval performances. 1 Introduction The effectiveness of multi media or hyper text systems depends on the ability to accurately retrieve images from a large database on the basis of contents [1, 2, 3, 4, 5]. The crucial ingredient of the underlying image database management system is an accurate way of indexing according not only to the frequency of salient image features but also to meaningful relational structure between the features. Ideally, such content based indices should be capable of ....

....the size of neighbourhood graph. Finally, we investigate the sensitivity of the representation to the addition of noise during the segmentation process of the image. 4. 1 Database Indexing Our first experiment is concerned with the retrieval of the n image that resemble most closely a query image [1, 2, 3, 4, 5]. Here, two query images (see figure 3) will be used to index into the database. However, the expected retrieval is very different whether we consider the Digital Map or the More Beef trademark. For both the following experiments the histograms where (a) The Digital Map (b) The More Beef ....

R. W. Picard, "Light-years from lena: Video and image libraries of the future", ICIP 95, vol. 1, pp. 310-- 313, 1995.

....from a segmentation of the image. The most severe problem with segmentation based approaches is that the decision as to whether a feature in an image is important is made at the time of data population. The segmentation is static, and subsequent queries do not have any influence on it. Picard [3] suggests pre computing multiple segmentations for an image, each reflecting a different aspect of the image. However, even when it is possible to segment an image in different meaningful ways, there is no way of automatically generating a set of segmentations that encompass all possible ....

R. W. Picard. Light-Years from Lena: Video And Image Libraries of the Future. In Proc. Int. Conf. on Image Proc., pages 310 -- 313, Washington D.C., USA, October 1995.

....in the previous section. The data base consists 28 infra red aerial images and over 1000 trademarks and logos. Our first experiment aims to demonstrate the retrieval power and accuracy of the proposed methodology. The basic aim of such systems is to retrieve from a large data base of images [9, 10, 14, 11], the n images that resemble most closely a query image. Here, the query image (see figure 3) is a Digital Map. This Digital Map corresponds to the area depicted on some of the infra red aerial images. The second experiment aims to investigate the sensitivity of recall to the line segmentation ....

R. Picard. Light-years from lena: Video and image libraries of the future. ICIP'95, 1:310--313, 1995.

....to the storage system and applications should be developed, such that even in huge volumes of information specific pieces can be located easily. Existing navigation systems are all based on prior annotation of the information or search processes on textual information. It is widely recognized [2 9] that there is a need for intelligent management and search methods in particular for the visual information in multimedia documents and in digital video and image libraries. The problems at hand are firstly that textual annotation can usually not be made beforehand because of the subjectivity and ....

R.W. Picard, "Light-years from Lena: Video and Image Libraries of the Future", Proc. of the IEEE Int. Conf. on Image Processing 1995, vol. 1, pp. 310-313, Washington DC, USA, 1995.

....and entertainment appliances. Such appliances can, however, only become a reality if powerful methods are developed for content characterization, browsing, filtering, and sorting. While a significant effort is being currently developed in these areas for purposes of content based retrieval [4], most approaches rely on image descriptions which are either specific to some problem domains (such as retrieval of news [9] or faces [6] or of very low level (such as image color and texture [2] This limits their scope or and their capability to support meaningful interaction with users which ....

R. Picard. Light-years from Lena: Video and Image Libraries of the Future. In Proc. Int. Conf. Image Processing, October 1995, Washington DC, USA.

....image of interest, the system instantly updates the screen showing other images in the database most similar to the selected image. The problems of what models to use for image representation, and how to measure image similarity are challenging research problems for the image processing community [57]. Photobook, like the systems of [55] and [56] allows the user to select manually from a variety of models and associated feature combinations. As a research tool, Photobook assists in rapid benchmarking of new pattern recognition and computer vision algorithms. An example interaction with ....

R. W. Picard, "Light-years from Lena: Video and image libraries of the future," in IEEE Second Int. Conf. on Image Proc., (Washington, DC), Oct. 1995. To appear; Also appears as MIT Media Lab Perceptual Computing TR #339.

....role; for example, Lenat s CYC common sense reasoning system can take a request such as find someone wet, and return an image with a label such as man finishing a marathon [1] The computer vision solutions will work best if wisely integrated with solutions from these other domains. In [2] I overviewed the latest digital library research issues for the image processing community to address. The emphasis for that community is on finding models for simultaneous compression and content description, and improving measures of visual similarity for comparing images. However, the domains ....

....wise for them to watch what each other does to insure complementary efforts and avoid wasteful duplication of errors, especially in spatiotemporal segmentation and modeling where research seems to overlap the most. In this paper I will focus on important research problems that I did not discuss in [2], problems which are traditionally closer to the computer vision community namely, generating descriptions of image content, and vision systems that learn. Although my focus in the rest of this paper is on computer vision for digital libraries, much of the following discussion also applies to ....

R. W. Picard, "Light-years from Lena: Video and image libraries of the future," in IEEE Second Int. Conf. on Image Proc., (Washington, DC), Oct. 1995. To appear; Also appears as MIT Media Lab Perceptual Computing TR #339.

No context found.

R. W. Picard, Light-years from Lena: Video and Images Libraries of the Future, International Conference on Image Processing (ICIP'95), Washington, DC, USA, October, 1995.

No context found.

R.W. Picard, "Light-years from Lena: Video and Image Libraries of the Future," IEEE International Conference on Image Processing, Washington DC, Oct. 1995.

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