First and foremost, thanks to Diego Catalano for his created algorithm libraries which has benefited many developers and programmers.
For the Object Fidelity algorithm as posted here: Is there an algorithm to detect the differences between two images? , the float results for
- Number of Errors
- Number of Mean-Square Errors
- Number of Mean-Square Signal-to-Noise Errors,
how can I convert there errors to an accuracy percentage of the similarity extent between two images?
I have limited experience on image processing design and analysis and owing to this library can improve the functionality of a gallery comparison framework.
Gladly appreciate any advice or guidance. Thanks a lot!