Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Histogram intersection return wrong distance #156

Closed
FabRez opened this issue Apr 7, 2023 · 2 comments · Fixed by #158
Closed

Histogram intersection return wrong distance #156

FabRez opened this issue Apr 7, 2023 · 2 comments · Fixed by #158
Assignees

Comments

@FabRez
Copy link

FabRez commented Apr 7, 2023

Histogram intersection method returns $\rm 1 - H(I,M)$ instead of $\rm H(I,M)$.

From Swain and Ballard 1991, Histogram intersection is defined as :

$\rm H_{SW}(I,M) = \dfrac{\sum_i \min(I_i,M_i)}{\sum_i M_i}$

The class HistrogramIntersection, instead, calculates it as:

$\rm H_{FR}(I,M) = 1 - \dfrac{\sum_i \min(I_i,M_i)}{\sum_i M_i} = 1 - H_{SW}(I,M)$

This way the class calculates the "percentage" of non-intersection rather than that of the intersection of the two histograms.

@jaime-cespedes-sisniega
Copy link
Contributor

@FabRez Thanks for reporting this issue.

I have renamed that class to HINormalizedComplement to reflect that what you want to calculate is a distance between two distributions using the normalized complement of the histogram intersection of that distributions.

@FabRez
Copy link
Author

FabRez commented May 4, 2023

@jaime-cespedes-sisniega Thank you for fixing it. I am glad that this was useful

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants