Citation:
Belongie, Serge, Jitendra Malik, and Jan Puzicha. "Shape matching and object recognition using shape contexts." Pattern Analysis and Machine Intelligence, IEEE Transactions on 24.4 (2002): 509-522.
Publication Link
Summary:
This paper is very similar to the $p algorithm, in the fact that it tries to match every point in the candidate to some point in the template. We take the set of vectors originating from a point to all other sample points on a shape. These vectors express the configuration of the entire shape relative to the reference point.
Discussion:
The following library has Javascript implementations of Haudroff Distance and Shape Context. This was the motivation behind looking unto Shape Context as an alternative means of template matching.
https://github.com/kjkjava/Sketchy.js/tree/master
The shape descriptor of a point is represented as a relative distribution over positions, in histogram bins of log-polar space. The cost of this matching for a point is defined using the chi-square statistic:
We choose the matching that minimizes the total cost of this matching.
Belongie, Serge, Jitendra Malik, and Jan Puzicha. "Shape matching and object recognition using shape contexts." Pattern Analysis and Machine Intelligence, IEEE Transactions on 24.4 (2002): 509-522.
Publication Link
Summary:
This paper is very similar to the $p algorithm, in the fact that it tries to match every point in the candidate to some point in the template. We take the set of vectors originating from a point to all other sample points on a shape. These vectors express the configuration of the entire shape relative to the reference point.
Discussion:
The following library has Javascript implementations of Haudroff Distance and Shape Context. This was the motivation behind looking unto Shape Context as an alternative means of template matching.
https://github.com/kjkjava/Sketchy.js/tree/master
The shape descriptor of a point is represented as a relative distribution over positions, in histogram bins of log-polar space. The cost of this matching for a point is defined using the chi-square statistic:
We choose the matching that minimizes the total cost of this matching.
No comments:
Post a Comment