Parametric Correspondence and Chamfer Matching: Two New Techniques for Image Matching

By H. Barrow, Josh Tenenbaum, R. Bolles, H. C. Wolf. 1977

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Deep Level Sets: Implicit Surface Representations for 3D Shape Inference (2019, methodology): “Here we considered two separate metrics for evaluating shapes inferred by our network, the Intersection over Union (IoU), also known as the Jaccard Index [19], and the Chamfer distance [1].”
Learning to Reconstruct Shapes from Unseen Classes (2018, methodology): “We hence evaluate reconstruction quality using Chamfer distance (CD) [Barrow et al., 1977], defined as CD(S1, S2) = 1 |S1| ∑ x∈S1 min y∈S2 ‖x− y‖2 + 1 |S2| ∑ y∈S2 min x∈S1 ‖x− y‖2, (1) where S1 and S2 are sets of points sampled from surfaces of the 3D shape pair.”
Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling (2018, methodology): “Here, we conduct behavioral studies to calibrate IoU, Chamfer distance (CD) [2], and Earth Mover’s distance (EMD) [45] on how well they reflect human perception.”
Leaf segmentation in plant phenotyping: a collation study (2015, methodology): “The framework is motivated by the well-known Chamfer Matching (CM) [8], which aligns one object instance in an image with a given template.”
Estimating 3D hand pose from a cluttered image (2003, methodology): “The chamfer distance [2] is a well-known method to measure the distance between two edge images.”
Maintaining Natural Image Statistics with the Contextual Loss (2018, methodology): “(8) is equivalent to the Chamfer Distance [22], while the other implies it is equivalent to the Contextual loss of [19].”
A search engine for 3D models (2003, methodology): “This helps make our method robust to small variations in the positions of lines, as in Chamfer matching [Barrow et al. 1977] and Hausdorff matching [Huttenlocher et al. 1993].”
Learning to Infer and Execute 3D Shape Programs (2019, background): “Evaluation metrics include IoU, Chamfer distance (CD) (Barrow et al., 1977), and Earth Mover’s distance (EMD) (Rubner et al., 2000).”
Multi-leaf tracking from fluorescence plant videos (2014, methodology): “Most prior image alignment work, such as ASM [14], AAM [15], Chamfer matching [16] and its extensions [17– 21], concentrate on aligning a single object instance.”
Voting-based pose estimation for robotic assembly using a 3D sensor (2012, methodology): “Following the seminal paper on chamfer distance [5], there were several useful variants that incorporate edge orientation [6], [7] or employ hierarchical representation [8].”
3D Shape Perception from Monocular Vision, Touch, and Shape Priors (2018, unknown): “Our metric is the classic Chamfer distance (CD) [49], widely used in the computer graphics community for measuring shape similarity.”
PointCleanNet: Learning to Denoise and Remove Outliers from Dense Point Clouds (2019, background): “If we assume the ground truth point clouds have a regular distribution, the following Chamfer measure [FSG17, ADMG18], a variant of the Chamfer distance [BTBW77], measures both of these properties: Here N and M are the cardinalities of the cleaned ˜ P and ground truth P point clouds, respectively.”
Distance transformations in digital images (1986, background): “This DT, which is also found in [7], is described by the mask in Fig.”