Método para identificação de pelos utilizando imagens naturais

Abstract

The automatic image recognition for diagnostic of hirsutism is not a common practice. If there is suspected of some cause of the hirsutism types, laboratory tests are oriented to be performed. Many factors are taken considering for the diagnostic, for example, time and development of hirsutism, family history, ethnicity, medications and physical examination (hair distribution patterns) are important for the evaluation of excess hair. For physical examination, the modified Ferriman-Gallwey scale is used for evaluation of hairs. Thus, this work uses the modified Ferriman-Gallwey scale. This scale evaluates the hair distribution and generates a score to define the intensity of hirsutism, also provides as a model to propose the extraction of a region of interest through image processing. For the simulations, a set of images from the literature review were analyzed. From these images, we are looking for obtaining the extraction of an ROI, through image processing, it was in accordance with the Ferriman-Gallwey scale model. After the extraction, this work could identify the regions that concentrate the most presence of hair contributing to the early diagnostic of hirsutism.