Alopecia Pattern Detection in Males using Classical Machine Learning

Alopecia is a problem faced by many adults under a certain age, sometimes due to hereditary reasons and others due to mental health factors. Medical clinics have proven to be a great help, but unfortunately, Alopecia is detected in later stages due to the lack of action from both sides. For many such reasons adult life may seem exigent. This research study presents a Machine Learning and computer vision-based approach for identifying the level of alopecia a male is suffering through the detection of the type. The Daegu University dataset was compiled with a hair segemneation data set available on male hair images. The balding pattern features are extracted using an ORB detector and descriptor. The large dimensions of the feature vector were optimized using K-means clustering and PCA. The paper represents an analysis of the classification performance of different classifiers such as KNN and SVM (poly) which observed an accuracy of81% and 78% respectively for balding pattern detection.