Glaucoma is the second leading cause of blindness worldwide. Often the optic nerve head (ONH) glaucomatous damage and ONH changes occur prior to loss of visual function and are observable "in vivo". Thus, digital image analysis is a promising choice for detecting the onset and/or progression of glaucoma. In this work, we present a new framework for detecting glaucomatous changes in the ONH using the method of proper orthogonal decomposition (POD). A baseline topograph subspace is constructed using POD for each eye to describe the ONH structure of the eye at a reference/baseline condition. The bases that form the baseline subspace capture the topograph measurement variability and any inherent structure variability of the ONH at baseline. Any glaucomatous changes in the ONH of an eye present during a follow-up exam are estimated by comparing the ONH topograph acquired from the follow-up exam with its baseline topograph subspace representation. Image correspondence measures of correlation, Euclidean distance, and image Euclidean distance (IMED) are used to quantify the ONH changes. An ONH topograph library built from the Louisiana State University experimental glaucoma study is used to demonstrate the performance.