Research On Key Technologies Of Image Segmentation Contextual Reasoning Based On Graph Convolutional Networks

Visual impairment includes some or all obstacles to the function of visual organs caused by congenital or acquired reasons.A series of situations from amblyopia or visual field defect to total blindness,from color weakness to color blindness belong to visual impairment.There are a large number of people with visual impairment,and as a group of amblyopia in the society,they deserve the attention and care of the society.Especially in the modern society with the rapid development of science and technology,the emergence of a large number of intelligent devices has enriched people’s social life.Visually impaired groups should deservedly enjoy the benefits brought by science and technology.Therefore,the development of relevant intelligent auxiliary devices has been highly valued.At present,the auxiliary equipment for visual impairment is not perfect,and most of the focus is on the travel obstacle avoidance or navigation of fully blind people.This topic fully considers the needs of amblyopia groups,vision impaired people and color blind people,and applies artificial intelligence algorithm and digital image processing technology to the auxiliary equipment for visual impairment.This topic selects a wearable augmented reality glasses and carries out secondary development on it.It uses lightweight target detection and human posture detection methods,and focuses on reminding the amblyopic groups of the detected contents,so as to assist their life and travel;The image recoloring technology is used to correct the picture seen by the color blind people,so as to help the color blind people better obtain the information in the image;Use the function of regional field of view transfer to place the picture in the visible field of view to help people with visual field defects see the complete life scene.At the same time,a corresponding APP has been developed for visually impaired users to choose the functions they need.