Computer libraries used in analysis of dermatological images with computational vision: a literature review

Abstract

Skin image analysis plays a fundamental role in the field of dermatology, as it enables early and accurate detection of various skin conditions. However, this process faces significant challenges due to the variability of characteristics present in skin lesions, such as textures, tones, and the existence of villi on the contour. In this article, a systematic review of literature on computer libraries used in the analysis of dermatological images with computer vision is presented. This research is based on the PRISMA statement and scientific databases: SCOPUS and IEEE Xplore for searching and aims to identify a wide variety of computer libraries and skin lesions. The results showed 7 libraries and 21 dermatological lesions, which contribute to a more precise analysis and a more reliable clinical diagnosis for the timely detection of skin disorders. In conclusion, this research highlights computer libraries that have a significant impact on improving clinical diagnosis, which is key to the development of effective solutions for people's health.