SKIN TYPE CLASSIFIER

Abstract

Variations in skin diseases largely fall into the subset of skin types. Classification of Skin Types plays an essential part in the classification of skin diseases. The paper proposes an end to-end machine learning model used to classify skin type as dry, oily, or normal using Python Libraries. Skin type classification is done using Gray-Level Co-Occurrence Matrix (GLCM) for texture identification and Light Gradient Boosting Machine (LGBM). Red Green Blue (RGB) images are first resized and then converted into Gray-Scale images and then features like energy, correlation, dissimilarity, homogeneity, contrast, and entropy are extracted from a given image using GLCM. Such features give us a basic understanding of the type and texture of the image. The GLCM of the image is calculated by changing the distance and angle of the pixels within the image in a regular interval. These features after extraction are fed to the LGBM algorithm. The trained model is then deployed using Flask. Key Words: features extraction, GLCM, LGBM, skin type