Is a Precise Distortion Estimation Needed for Computer Aided Celiac Disease Diagnosis?

Abstract

In computer aided celiac disease diagnosis, endoscopes with wide-angle lenses are deployed which induce significant lens distortions. This work investigates an approach to automatize the estimation of the lens distortion, without a previous camera calibration. Knowing the discriminative power of all sensible distortion configurations, the model parameters are estimated. As the achieved parameters are not highly precise, moreover, we investigate the effect of approximative distortion correction on the classification accuracy. Particularly, we identify one simple but especially for certain features highly effective approximative distortion model.

Impact of image segmentation techniques on celiac disease classification using scale invariant texture descriptors for standard flexible endoscopic systems

“…This descriptor[30] recognizes a texture by calculating the difference between the average gray pixel density difference level represented by binary patterns. This method is frequently used in the classification of CDs because spatial image representations are very promising and have been shown to be very robust and accurate[4,7,21,[31][32][33][34][35] in the literature. The input image condition is real and nonsparse, an M -by -N 2-D grayscale image.…”