This document presents a new algorithm for mobile vision-based localization of skewed nutrition labels on grocery packages, achieving high specificity by minimizing false positives. The algorithm employs edge, line, and corner detection methods, accommodating labels skewed by up to 40 degrees from vertical, and is implemented on a Google Nexus 7 smartphone. Performance evaluation on a dataset of 378 images shows a true positive rate of 42% and emphasizes the importance of rapid processing to enhance user engagement in nutrition management.