The aging of infrastructure has become a significant social issue, drawing attention to distributed optical fiber sensors for monitoring structural defects. Among these, Brillouin optical correlation domain reflectometry (BOCDR), which uses continuous light correlation control for spatial resolution, has been advanced to improve performance. BOCDR measures the distribution of Brillouin gain spectra in the fiber under test, allowing the extraction of Brillouin frequency shift distributions, which can be related to strain and temperature. However, traditional slope-assisted (SA-) BOCDR techniques have limited accuracy in real-time operation due to the slow frequency sweep of the electrical spectrum analyzer. To address this, double-slope-assisted (DSA-) BOCDR has been proposed, utilizing power ratios of two different frequencies for enhanced loss point detection accuracy. In this study, we experimentally demonstrated improved localization accuracy of loss points using DSA-BOCDR. We applied two distinct strains and a bending loss at different positions on a silica single-mode fiber and measured power variations at two frequencies (10.815 and 10.865 GHz). The results showed that the DSA-BOCDR could accurately detect strain and loss locations, with the power ratio method yielding sharper indications of loss compared to single-frequency measurements. The differential analysis further quantified the loss range, showing higher positional accuracy for loss detection using the power ratio. These findings confirm that using distinct power ratios of two frequencies in DSA-BOCDR significantly enhances the precision of loss point detection, offering a promising advancement for real-time structural health monitoring in aging infrastructures.