Image processing inventions span a broad range of technologies - military (IR), medical (CT, MRI, PET, ..), AI (facial recognition), and the list goes on. Depending on the examiner handling your image processing invention, there is a wide spectrum of how Section 101 might be applied. If you get an examiner aiming to reject the application, look out for Section 101 rejections and reliance on the ever-popular Electric Power decision. Further, there have been PTAB decisions also seeming to undercut the patentability of image processing as evidence by the Hitachi Medical case.
However, all is not lost and at least some panels at the PTAB recognize the patentabilty of image processing that is achieved with pure data manipulation as illustrated by the more recent decision in the CGGVeritas Services case. In this case, the claimed method steps on appeal are:
The PTAB walks through the relevant case law, including citation to SAP, McRo, Finjan, but also includes a cite to Electric Power (although only for the principle that step one of the inquiry is sometimes dispositive). But what seems to matter here is the claims do not simply recite a desired result (e.g., a desired image), and instead include specific features that are used to achieve the desired result. From the decision:
The PTAB then goes on to compare the case to Thales Visionix (the physical tracking system) and since the claims here were focused on a physical-realm improvement, the claims were patent eligible (and thus there was no need to consider step two of the Alice inquiry).
So, if you have an image processing invention being rejected as too abstract, you may want to review the Veritas decision as part of your strategy to address the rejection and possibly prepare the case for appeal.