The term chamfer distance was found in the process of reading paper, so I checked the online information to summarize.
A distance transform for images, commonly used in shaped based object detection. For a binary image with feature points and non-feature points, this distance transformation is to find the distance from each point to the nearest feature point.
Chamfer Distance
Matching process
Match two images: one of them calculates the Chamfer distance transform, superimposes the feature points of the other image on the DT, and calculates the average of the DT values corresponding to the feature points, then the distance between the curve and the image can be superimposed by these It is calculated by some mean value of DT at the point, such as root mean square (rms).
