ACS/IEEE 2005 International Conference on Computer Systems and Applications (AICCSA'05)
Enhanced visual evaluation of feature extractors for image mining
Cairo, Egypt
January 03-January 06
ISBN: 0-7803-8735-X
M. Malki, Comput. Sci. Dept., Sao Paulo Univ., Sao Carlos, Brazil
A.J.M. Traina, Comput. Sci. Dept., Sao Paulo Univ., Sao Carlos, Brazil
A.J.M. Traina, Comput. Sci. Dept., Sao Paulo Univ., Sao Carlos, Brazil
Summary form only given. This paper introduces a novel approach to evaluate, timely and effectively, the suitability of new image feature extraction techniques concerning similarity queries using CBIR systems. The proposed approach is based on two measurements derived from spatial properties intuitively and naturally perceived in spatial domains, and that can also be verified in multidimensional spaces. To bear out our proposal, we show that the insights obtained by the proposed measurements comply with the well-known analysis methods based on the precision and recall approach.
Citation:
M. Malki, A.J.M. Traina, A.J.M. Traina, "Enhanced visual evaluation of feature extractors for image mining," aiccsa, pp.45-I, ACS/IEEE 2005 International Conference on Computer Systems and Applications (AICCSA'05), 2005