What is the best compression method for a 2500 ppi image?

There is no difference in algorithms that depend on image resolution. Resolution is just a parameter that maps the number of pixels in the image to physical units of measure - typically it is called "300 dpi" (meaning 300 ppi pixels per inch and not dots- per-inch ) into "images for printing."

2500 PPI is a resolution lady - possibly used in colored slide films. Even the craziest LCD displays are below or around 500 PPI resolution.

So if you have an image to appear in 2 x 1 inch at 2500 PPI it will still be an image of 5000x2500px -e it's that number of pixels that you have to compress. The resolution is 2500 PPI, or 1 pixel per meter (in which case the same image would extend by 5Km of landscape), does not differ in any way the algorithmic treatment on the pixels. The resolution is a single few-byte field in the final file, which serves as the "tip" of the final size for some of the preview / print programs.

I have now seen that this WSQ standard was developed to handle fingerprint image storage - hence we have an idea of your domain and the types of images you need to store. It may be that the WSQ is great for storing the details of a fingerprint that appears on the 500PPI scale - and misses on others - but that has to do with the type of image and type of details specific to your domain. If WSQ is not good at other resolutions, and you want something lossless, use the same PNG files - which is good enough for generic images - if they are too big for your application, a format like Google WebPP may work better.