Artificial Vision System for Physical Activity Practice for People with Visual Impairment

Visual impairment is a congenital or developed condition, sometimes it tends to affect from childhood to old age, it can be permanent or treated with a previous diagnosis, sometimes it is corrected with special lenses, and in other cases, expensive surgeries are required. Studies have shown a decrease in physical activity, an increase in sedentary lifestyles, and the probability of developing diseases such as obesity and overweight as a result. The World Health Organization recommends physical activity to maintain good health for people with this condition. This work presents the development of a system for correcting the orientation of visually impaired people while practicing walking as a physical activity, using image processing and implementing computer vision. The system provides feedback to the user with vibrations and sounds to correct its trajectory, using existing technologies and algorithms such as the Hough Transform. To prove its performance, the system was tested on people in simulated blindness conditions on athletic tracks. The subjects walked 100 meters in a straight line using the device. The results show that the device can keep the user walking in a straight line and correct directional changes to avoid leaving the lanes of an athletic track.