Due to the increasing prevalence of sensitive skin, it is crucial to effectively evaluate the skin barrier function. As the gold standard for assessing the skin barrier, Transepidermal Water Loss (TEWL) has been limited in its widespread use due to its high measurement costs and stringent environmental requirements. This study proposes dermatoscopic image texture features for the assessment of the sensitive skin barrier, Specifically, we extracted Gray Level Co-occurrence Matrix (GLCM) and Local Binary Pattern (LBP) features after pre-processing the collected dermatoscopic images, which are expected to replace TEWL. Experimental results from model mice indicate a correlation between these texture features and TEWL values, suggesting that they can be regarded as effective indicators for assessing the barrier function of sensitive skin.