Dermoscopy image automatic segmentation method based on full convolutional neural network

Abstract

The present invention provides a dermoscopy image automatic segmentation method based on a full convolutional neural network. The method comprises the following four steps: 1: obtaining of dermoscopy images and true value graphs; 2: performing structure design of a full convolutional neural network; 3: performing design of feature fusion and a per-pixel segmentation method; and 4: performing network training and segmentation. According to the steps mentioned above, an end-to-end depth convolutional neural network is obtained through training to perform accurate segmentation of the dermoscopy images and allow a small area skin lesion area to be effective so as to solve actual problems that the skin lesion area segmentation is not good to influence the subsequent diagnosis accuracy in the dermatology computer auxiliary diagnosis system.

Description

(1) Technical field: [0001] The invention relates to a method for automatically segmenting dermoscopic images based on a fully convolutional neural network, and belongs to the technical fields of image processing and machine learning. (two) background technology: [0002] As the first line of defense against the invasion of external pathogens, the skin plays a vital role in human health. However, with changes in the environment and human diet, various skin diseases affect human life, and skin cancer is a threat to human life. Dermoscopy is a non-invasive microscopic image analysis technique for observing the microstructure and pigments below the surface of living skin, which is of great significance for the clinical diagnosis of skin diseases. [0003] When clinicians use the naked eye method to diagnose skin diseases through dermoscopy, they often rely on their own experience and subjective visual evaluation, resulting in low diagnostic accuracy and poor repeatability. Th...

AI Technical Summary

Problems solved by technology

Segmentation of dermoscopic images is very challenging due to various conditions such as low contrast, blurred lesion edges, variable color texture, and hair noise in dermoscopic images.

Images