Distinctive Image Features from Scale-Invariant Keypoints
Citations
106 citations
Cites methods from "Distinctive Image Features from Sca..."
...Using the standard parameters proposed by Lowe [5], SIFT features are extracted from an expanded window around the license plate region from image I, and matched to other features extracted from image J, using the nearest neighbor distance ratio matching strategy described by Mikolajczyk [34]....
[...]
...We have used the SIFT (ScaleInvariant Feature Transform) features proposed by Lowe [5]....
[...]
...To cope with large displacements, from vehicles moving at high speeds, an initial motion estimation is performed by matching features extracted by the ScaleInvariant Feature Transform [5]....
[...]
106 citations
Cites methods from "Distinctive Image Features from Sca..."
...All of the shallow methods are tested with hand-crafted features (SIFT) [49] and deep features (VGGNet)....
[...]
...For shallow methods, we use both hand-crafted (SIFT) and deep features (VGGNet) as inputs for comparisons....
[...]
106 citations
Cites background from "Distinctive Image Features from Sca..."
...2) The visual and semantic spaces are completely different: image descriptors are generally in a continuous space (e.g., SIFT [7] is typically represented as a 128-...
[...]
...SIFT [7] is typically represented as a 128-D real-valued vector), whereas semantic attributes are denoted in a discrete space....
[...]
...In this method, the tree is obtained by hierarchical K-mean clustering of 1 million SIFT descriptors extracted from images randomly crawled from the internet....
[...]
...Although there are similarities between image and text information retrieval, with low-level features quantized into visual words and inverted file indexing applied to index images via visual words [7], two major challenges still exist for image retrieval: algorithm and system design....
[...]
105 citations
Cites background from "Distinctive Image Features from Sca..."
...To date, feature matching [4], [5], [6] is the correspondence solution of choice for many computer vision systems....
[...]
105 citations
Cites background from "Distinctive Image Features from Sca..."
...One contribution of that paper was to demonstrate the suitability of local detectors (scale, rotation, and affine invariant Harris and Laplace) and descriptors (SIFT [27], spin images [28] and RIFT [29]) for these tasks....
[...]
References
46,906 citations
16,989 citations
13,993 citations
7,057 citations
3,422 citations