Image persistence with sparse distance matrices

[mglisse]

Hello,
we wanted to compute the persistence diagram of the image between 2 filtrations, and the branch https://github.com/Ripser/ripser/tree/image-persistence-simple looked promising. However, it looks like it is based on an old version of Ripser before sparse distance matrices were introduced, and our (very sparse) matrices are too large for a dense representation.
Do you think it is mostly a matter of merging the 2 branches (which may be much more work than git merge but should remain relatively straightforward), or is there something in the image algorithm that is likely to behave badly with sparse distance matrices?