I work for a Motorola biometrics unit. Our core business is fingerprint images: extracting minutia, cleaning them up, etc. I work in the infrastructure part of things, so I don't interact with image processing on a regular basis. A few months back a colleague of mine was tasked with finding a good library to base our future image processing development. After some review of the available libraries, he recommended GIL. His main reasons was the very good idea of view/image separation; the establishing a lightweight bounds checker/accessor to the image's pixels. I decided to review the library myself and found it to be very intuitive. As I mentioned earlier I work more in the business process/infrastructure side of things, so this was my first serious interaction with image processing. As a learning project, I decided to implement a few functions to get acquainted with the concepts of both GIL and the image processing. Within a week and a half of reading wikipedia and looking at GIL, I ended up implementing a convolution view. I enjoyed working with and learning from GIL; it is definitely a well thought out library. We will be using it for our needs. Tervel Atanassov