The present work seeks to address a recurring issue in swine production that, despite technological and scientific advances, remains frequently overlooked—largely due to producers’ limited technical knowledge. This study proposes the development of an automated method for detecting and classifying swine hoof lesions, with the aim of enabling early identification and preventing consequences such as reduced herd longevity and reproductive disorders. Ultimately, the goal is to integrate the detection and classification model into an embedded system implemented as a mobile application, thereby making lesion monitoring and classification more practical, efficient, and accessible.