Multi-scale phase-based local features
Gustavo Carneiro,Allan D. Jepson +1 more
- 18 Jun 2003
- Vol. 1, pp 736-743
TL;DR: The results show that the phase-based local features lead to better performance than the other two approaches when dealing with common illumination changes, 2D rotation, and sub-pixel translation.
read more
Abstract: Local feature methods suitable for image feature based object recognition and for the estimation of motion and structure are composed of two steps, namely the 'where' and 'what' steps. The 'where' step (e.g., interest point detector) must select image points that are robustly localizable under common image deformations and whose neighborhoods are relatively informative. The 'what' step (e.g., local feature extractor) then provides a representation of the image neighborhood that is semi-invariant to image deformations, but distinctive enough to provide model identification. We present a quantitative evaluation of both the 'where' and the 'what' steps for three recent local feature methods: a) phase-based local features (Carneiro and Jepson, 2002), b) differential invariants (Schmid and Mohr, 1997), and c) the scale invariant feature transform (SIFT) (Lowe, 1999). Moreover, in order to make the phase-based approach more comparable to the other two approaches, we also introduce a new form of multi-scale interest point detector to be used for its 'where' step. The results show that the phase-based local features lead to better performance than the other two approaches when dealing with common illumination changes, 2D rotation, and sub-pixel translation. On the other hand, the phase-based local features are somewhat more sensitive to scale and large shear changes than the other two methods. Finally, we demonstrate the viability of the phase-based local feature in a simple object recognition system.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Figures
Figure 3: All image deformations with a false positive rate fixed at 0.01 and computing the detection rate for varying amount of change. Here, the phase-based, differential invariant and SIFT features are represented by the solid, dotted and dashed lines, respectively. The vertical axis represents detection rate and the horizontal axis shows the amount of variation 
Figure 6: Top left image: user selected “baking soda box” model. Sequence: searching the model over a series of cluttered images containing the model at different poses and partially occluded. The light points inside the distorted rectangles represent the interest points used for the best similarity match. 
Figure 1: Comparison between the scale robust interest point detector described above (solid line), and the interest point detectors Harris-Laplacian (dotted curve) and difference-of-Gaussian (dashed curve). 
Figure 2: Configuration of local descriptor for . 
Figure 4: Top left image: segment of the image selected by the user to define the “boy’s left eye” model. Other images: recognizing the model over a sequence of 100 images (only 5 are shown). The light points inside the distorted rectangles represent the interest points used for the best similarity match. 
Figure 5: Top left image: segment of the image selected by the user to define the “tetley box” model. Remaining images: the light points inside the rectangles represent the interest points used for the best similarity match.
Citations
SURF: speeded up robust features
Herbert Bay,Tinne Tuytelaars,Luc Van Gool +2 more
- 07 May 2006
TL;DR: A novel scale- and rotation-invariant interest point detector and descriptor, coined SURF (Speeded Up Robust Features), which approximates or even outperforms previously proposed schemes with respect to repeatability, distinctiveness, and robustness, yet can be computed and compared much faster.
Speeded-Up Robust Features (SURF)
TL;DR: A novel scale- and rotation-invariant detector and descriptor, coined SURF (Speeded-Up Robust Features), which approximates or even outperforms previously proposed schemes with respect to repeatability, distinctiveness, and robustness, yet can be computed and compared much faster.
14.9K
Description of interest regions with local binary patterns
TL;DR: A new texture feature called center-symmetric local binary pattern (CS-LBP) is introduced that is a modified version of the well-known localbinary pattern (LBP), and is computationally simpler than the SIFT.
1.3K
Multi-image matching using multi-scale oriented patches
Matthew Brown,Richard Szeliski,Simon Winder +2 more
- 20 Jun 2005
TL;DR: This paper describes a novel multi-view matching framework based on a new type of invariant feature that is used in an automatic 2D panorama stitcher that has been extensively tested on hundreds of sample inputs.
Evaluation of Interest Point Detectors and Feature Descriptors for Visual Tracking
TL;DR: This work presents a carefully designed dataset of video sequences of planar textures with ground truth, which includes various geometric changes, lighting conditions, and levels of motion blur, and presents a comprehensive quantitative evaluation of detector-descriptor-based visual camera tracking based on this testbed.
479
References
Object recognition from local scale-invariant features
David G. Lowe
- 20 Sep 1999
TL;DR: Experimental results show that robust object recognition can be achieved in cluttered partially occluded images with a computation time of under 2 seconds.
A Combined Corner and Edge Detector
Chris Harris,Mike Stephens +1 more
- 01 Jan 1988
TL;DR: The problem the authors are addressing in Alvey Project MMI149 is that of using computer vision to understand the unconstrained 3D world, in which the viewed scenes will in general contain too wide a diversity of objects for topdown recognition techniques to work.
The design and use of steerable filters
TL;DR: The authors present an efficient architecture to synthesize filters of arbitrary orientations from linear combinations of basis filters, allowing one to adaptively steer a filter to any orientation, and to determine analytically the filter output as a function of orientation.
Evaluation of Interest Point Detectors
TL;DR: Two evaluation criteria for interest points' repeatability rate and information content are introduced and different interest point detectors are compared using these two criteria.
Indexing based on scale invariant interest points
Krystian Mikolajczyk,Cordelia Schmid +1 more
- 07 Jul 2001
TL;DR: In this article, scale invariant interest points are used for image indexing, which is based on two recent results on scale space: (1) interest points can be adapted to scale and give repeatable results (geometrically stable).
Related Papers (5)
Chris Harris,Mike Stephens +1 more
- 01 Jan 1988
David G. Lowe
- 20 Sep 1999
Yan Ke,Rahul Sukthankar +1 more
- 27 Jun 2004