Artificial intelligence-based skin erythema identification method and device, medium and terminal

The invention provides a skin erythema recognition method and device based on artificial intelligence, a medium and a terminal, and the method comprises the steps: collecting a historical skin erythema image, and carrying out the preprocessing and image enhancement processing, so as to obtain a hist...

Full description

Saved in:
Bibliographic Details
Main Authors SUN YI, FAN XINDONG, SU LIXIN, CAI REN, JIANG CHUNYU, WANG ZHENFENG
Format Patent
LanguageChinese
English
Published 23.06.2023
Subjects
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
DIAGNOSIS
HUMAN NECESSITIES
HYGIENE
IDENTIFICATION
IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
MEDICAL OR VETERINARY SCIENCE
PHYSICS
SURGERY
Online AccessGet full text

Cover

More Information
Summary:The invention provides a skin erythema recognition method and device based on artificial intelligence, a medium and a terminal, and the method comprises the steps: collecting a historical skin erythema image, and carrying out the preprocessing and image enhancement processing, so as to obtain a historical enhanced skin erythema image; carrying out image segmentation on the ROI image by adopting a segmentation algorithm to obtain an ROI image; extracting erythema features in the ROI image, converting the erythema features into recognizable vectors, performing classification and recognition on the recognizable vectors based on a classification algorithm, and training a skin erythema recognition model based on machine learning; and identifying a to-be-detected skin erythema image by using the trained skin erythema identification model to obtain a corresponding erythema feature identification result, and performing analysis in combination with the evaluation factors to obtain a corresponding hemangioma and vascul
Bibliography:Application Number: CN202211492682