You have no way to know. The IP tracking is performed by the server that serves you the image. The server necessarily knows that it served some image file, but you have no way -- as a client -- to know whether the server does anything meaningful with that information, or whether the image load could be traced back to you personally.
If it's a mass email, the image URL would probably have your email address in it (or, more likely, some randomly-generated unique ID that maps to your email in a server-side database). That's so the server knows whose IP they're getting with each image load. (e.g., "Okay,
126.96.36.199 just loaded the URL
/images/35fd76a74/track.png... from our records, we sent that URL to
email@example.com.) However, if an information gathering effort is focused on only you (or the service is very clever in how it encodes unique IDs in their URL), the URL could look perfectly mundane.
In order to decide if an image is used to track your IP address, you'd have to decide whether the server associates the loaded image uniquely back to your email address. Since you don't run the server, this is generally not possible. If some image is known to be widely used (e.g., just a
/static/img/header.png that the sender puts in every single one of their emails to many recipients, without any kind of unique ID in the URL) then it's unlikely that that image could be used for tracking: too many people load that same image for the originating IPs of its requests to be uniquely linked back to a particular recipients.
If you fetch the images through an anonymizing proxy, you can hide your IP address. However, a proxy cannot hide the fact that the image was requested at all. The server would still know that
/images/35fd76a74/track.png was loaded (and that that unique ID links back to your email address), but they simply don't know what your real IP is.