Lopresti D. and Tomkins A. "Temporal domain matching of hand-drawn pictorial queries". In Proc. of the Seventh Conf of The Ind. Graphonomics Society, pp. 98-99. August 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....experiment is hand written text. This approach has the drawback that it treats the same hand drawings with different stroke orders as a poor match. Besides, the algorithm is sensitive to different writing styles. In order to make the system less sensitive to the stroke order, Lopresti and Tomkins [4] [5] proposed to match the strings block by block. Under the modified scheme, the string is divided into blocks which are used as the basic unit for matching between the query and the database. This modification allows the matching of the hand drawings with arbitrary drawing order of the strokes. ....

Lopresti D. and Tomkins A. "Temporal domain matching of hand-drawn pictorial queries". In Proc. of the Seventh Conf of The Ind. Graphonomics Society, pp. 98-99. August 1995.

....of Invariant Patches for Indexing into Pictorial Databases Pictorial databases and digital libraries have experienced tremendous growth recently. To be able to access the vast amount of pictorial information now available, many content based query and indexing systems have been introduced [12, 13, 14, 15, 16, 17]. Most of the current pictorial database indexing schemes are appearance rather than geometry based. Geometry based indexing allows rapid queries to be performed on large pictorial databases containing images with complex, rich shape structure. Figure 1 illustrates various modules involved in a ....

D. Lopresti and A. Tomkins, "Temporal-Domain Matching of Hand-Drawn Pictorial Queries," in Handwriting and drawing Research: Basic and Applied Issues (K. Simner, C. Leedham, and A. Thomassen, eds.), pp. 387--401, IOS Press, 1996.

....complexity of the algorithms is O(m 2 n) or O(m 2 n 2 ) de17 pending on the particular variant. Hence, although they are polynomial, the order of growth is sufficiently large that pruning of the search space is required for large problems. In our experimental efforts (see, for example, LT95] we have found that the algorithms below provide a general framework for block matching. Domain knowledge can then be incorporated to improve performance by pruning large subtress of computation when possible. 7 Acknowledgements The genetic sequence dot matrix plot (Figure 2) was generated ....

D. Lopresti and A. Tomkins. Temporal domain matching of hand-drawn pictorial queries. In Proceedings of the Seventh Biennial Conference of the International Graphonomics Society, pages 98--99, London, Ontario, 1995. Chapterlength version submitted for publication.

....developed for matching textual ink are not flexible enough to capture these kinds of block motion. In a recent paper, Lopresti and Tomkins introduced a family of algorithms for block string matching that address this issue [5] Preliminary results, based on a small experiment, seemed promising [6]. In the present paper, we give a more detailed analysis of the problem of matching hand drawn sketches. We describe a hierarchical approach that exploits the temporal nature of electronic ink, and evaluate it using a larger dataset. Figure 1 presents a conceptual overview. Sketch matching can ....

D. Lopresti and A. Tomkins. Temporal domain matching of hand-drawn pictorial queries. In Proc. Seventh Conf. Int. Graphonomics Society, pages 98--99, Aug. 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