Recognition of Impediments for Visually Impaired Using Open CV

Eye sight is one among the most vital senses for survival. Millions of individuals throughout the world suffer from vision impairment. People who are blind or visually handicapped have a more difficult time doing their regular duties. Even if they don't want to, the majority of them are dependents for assistance. A handful of technologies for visually impaired people's aid have been created. Because of its cost and accessibility, machine eyesight-based remedies are appearing as one among many effective technologies being used to help the blind. The suggested system attempts to provide a wearable visual aid that would allow visually impaired persons to move autonomously without the need for external support by utilising a live object detection arrangement. To detect hurdles uses a image processing (machine learning) mechanism and via the camera and notify visually impaired person about the obstacle and its position via voice output. On the Open CV platform, python programming language with Raspberry Pi 3 model - B is used to develop artificial eyesight. Proposed Classification accuracy of 80% in reading text and 90% in detecting the object is observed.