Personal color diagnostic method and system based on machine learning and augmented reality

The present invention relates to a personal color diagnosing method based on machine learning and augmented reality. The personal color diagnosing method based on machine learning and augmented reality enables a user to intuitively determine a personal color by diagnosing the personal color based on a personal color estimation model and generating and providing augmented reality content by corresponding to a diagnosis result. According to the present invention, the personal color diagnosing method based on machine learning and augmented reality includes: a step of receiving user information including at least one among sex, age, and a brand preference of the user from the user, receiving a user image including a face of the user from the user, and transmitting the user information and the user image, which are inputted, to a diagnosing server, by a user terminal; a step of preprocessing the received user image and determining classification information including at least one among a skin tone, an eye color, and a hair color of the user from the user image by analyzing the user image preprocessed, by the diagnosing server; a step of storing user data by combining the classification information extracted from the user image and the user information received from the user terminal by the diagnosing server and diagnosing the personal color in the user data according to the personal color estimation model by the diagnosing server; and a step of generating a recommended cosmetic list to be recommended to the user based on the deduced personal color and transmitting the recommended cosmetic list to the user terminal, by the diagnosing server. 본 발명은 머신러닝 및 증강현실 기반의 퍼스널 컬러 진단 방법에 관한 것이다. 사용자단말이 사용자의 성별, 나이, 선호브랜드 중 적어도 하나를 포함하는 사용자정보를 사용자로부터 입력받고, 사용자로부터 사용자의 얼굴을 포함하는 사용자이미지를 입력받고, 입력받은 사용자정보와 사용자이미지를 진단서버에 전송하는 단계와, 진단서버가 수신된 사용자이미지에 대해 전처리를 수행하고, 전처리된 사용자이미지를 분석하여 사용자이미지로부터 사용자의 피부톤, 사용자의 눈동자색상, 머리카락 색상 중 적어도 하나를 포함하는 분류정보를 판별하는 단계와, 진단서버가 사용자단말로부터 수신된 사용자정보와 사용자이미지로부터 추출된 분류정보를 통합하여 사용자데이터를 저장하고, 진단서버가 퍼스널컬러예측모델에 따라 사용자데이터에서 퍼스널컬러를 진단하는 단계와, 진단서버가 도출된 퍼스널컬러에 기초하여 사용자에게 추천할 추천화장품리스트를 생성하고, 사용자단말에 추천화장품리스트를 송신하는 단계를 포함한다.