Research On Image Classification Method Of Diabetic Retinopathy Based On Meachine Learning

In recent years,Deep Learning technology has been widely applied to people's daily lives since the increasing maturity of Artificial Intelligence technology.Many researchers pay attention to the medical intelligent.Using artificial intelligence to help doctors diagnose is an important branch of medical intelligent.As an important basis for the diagnosis of eye diseases,retinal fundus pictures contain many subtle information about eye diseases.Accurate analysis of the pictures can help patients get their treatment in time.It is a very meaningful topic,by using Artificial Intelligence to analyze eye pictures not only saving a lot of time to train an experienced ophthalmologist,but also reduces the number of mismatches between doctors and patients.In this topic,a grading method for diabetic retinopathy images based on machine learning algorithms is proposed.The classification model algorithm for the fundus picture database of two different data sets is designed and implemented.By introducing the morphological operation method in the segmentation algorithm,the training effect of the classification model under the small data set is improved.Small data sets are trained using support vector machines and integrated learning methods.The features extracted from the migration learning are trained by the support vector machine,and the combination of the neural network and the traditional machine learning method is realized.By introducing the RMSProp optimization algorithm into the convolutional neural network model,the problem of grading the fundus map with a large data set is solved.The experimental content will be used for the training of support vector machine and migration network for the small data set,and supplemented by the integrated learning model.Deep neural network model is used for the large data set for training,and it is divided into unprocessed data set after processing and unprocessed biased data set.This experiment aims to classify diabetic retinopathy by training in above models.