“…In Table 1, we have reported the precision of our embedded method against the precisions achieved by the approaches using only similarity function score [12] or only the Hausdorff distance [17]. We can observe that our method precision outperforms the state-of-art ones.…”
Section: Results Evaluation and Discussionmentioning
confidence: 67%
“…The latter ones could be based on Harris interest points [6], or are distribution-based descriptors such as scale invariant feature transform (SIFT) descriptors [7], shape contexts [8], or the speed-up robust features (SURF) [9] and more recently, DAISY [10]. The local descriptors could also be computed on a dense grid as it is the case for the histogram of oriented gradient (HOG) [11] or the packed dense interest points [12].…”
In this paper, we present a new approach for matching local descriptors such as Scale Invariant Feature Transform (SIFT) ones in order to recognize image objects quickly and reliably. The proposed method first computes the Hausdorff distance combined with the City-Block distance to match the two sets of the extracted keypoints from the goal and data images, respectively. Then, the matched points are involved into an embedded pairing process, leading to a double matching which is more discriminant for the object recognition purpose as demonstrated on real-world standard databases.
“…In Table 1, we have reported the precision of our embedded method against the precisions achieved by the approaches using only similarity function score [12] or only the Hausdorff distance [17]. We can observe that our method precision outperforms the state-of-art ones.…”
Section: Results Evaluation and Discussionmentioning
confidence: 67%
“…The latter ones could be based on Harris interest points [6], or are distribution-based descriptors such as scale invariant feature transform (SIFT) descriptors [7], shape contexts [8], or the speed-up robust features (SURF) [9] and more recently, DAISY [10]. The local descriptors could also be computed on a dense grid as it is the case for the histogram of oriented gradient (HOG) [11] or the packed dense interest points [12].…”
In this paper, we present a new approach for matching local descriptors such as Scale Invariant Feature Transform (SIFT) ones in order to recognize image objects quickly and reliably. The proposed method first computes the Hausdorff distance combined with the City-Block distance to match the two sets of the extracted keypoints from the goal and data images, respectively. Then, the matched points are involved into an embedded pairing process, leading to a double matching which is more discriminant for the object recognition purpose as demonstrated on real-world standard databases.
“…These local distribution-based descriptors are computed in surrounding regions of extracted keypoints by means of interest point detectors such as Hessian detector [17], Harris [18], Laplacian of Gaussian (LoG) [19], Difference of Gaussian (DoG) [12], etc. or by dense sampling such as in [20].…”
Abstract-Reliable and effective matching of visual descriptors is a key step for many vision applications, e.g. image retrieval. In this paper, we propose to integrate the Hausdorff distance matching together with our pairing algorithm, in order to obtain a robust while computationally efficient process of matching feature descriptors for image-to-image querying in standards datasets. For this purpose, Scale Invariant Feature Transform (SIFT) descriptors have been matched using our presented algorithm, followed by the computation of our related similarity measure. This approach has shown excellent performance in both retrieval accuracy and speed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.