To date, approximately 12 million people over the age of 40 in the United States are visually impaired, 3 million of whom are blind. Meanwhile, many people are blinded by accidents. Guide dogs can assist blind people in their daily life. However, the cost to breed, raise, train, and match a Guiding Eyes dog with a person with vision loss is estimated to be approximately $50,000. Another weakness is that the life expectancy of a human is 73.4 years around the world and the average working life of a guide dog is around eight years. As a result, people with visual impairments need more guide dogs throughout their lives. If our design is put into use, we can reduce the cost significantly. In this paper, we are going to build a smart seeing eye dog, which is a kind of wheeled Assistive Robotics that could guide the blind people walk on the road like but perform better than a real seeing eye dog. We started this paper from four aspects. They are the structure of the mechanical dog, TensorFlow-based image recognition, ROS-based motion simulation and human-computer interaction. Because of the cost of training an assistive dog is extremely high and the dogs are all red-green color-blind, which means that they cannot recognize the traffic lights and road sign. In our paper, these problems can be solved very well. Soon, between this mechanical dog paper may help blind people to better integrate into society.
Smart Seeing Eye Dog Wheeled Assistive Robotics
24.09.2021
3149330 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Shared Control Templates for Assistive Robotics
Deutsches Zentrum für Luft- und Raumfahrt (DLR) | 2020
|Employing Whole-Body Control in Assistive Robotics
Deutsches Zentrum für Luft- und Raumfahrt (DLR) | 2019
|Special Issue on Assistive and Rehabilitation Robotics
British Library Online Contents | 2017
|