Deep Learning and Fuzzy Decision Support System for visually impaired persons

This paper presents the concept of wearable guidance device using deep learning to support blind or visually impaired people. Using suggested wearable technology, this effort seeks to provide supplemental support to them with more flexible movements. The large volume of device data also assures that its users are safer. The suggested device employs an RGB camera to transform RGB photos and compute a level surface for recognizing obstacles and safe walking paths, deep learning is used. The proposed navigation system takes sensor data as input and gives the BVIP the necessary safety orientation. To assist BVIP in choosing a safe course, a fuzzy logic-based decision support system was built. The method has been tested on blinded and visually challenged individuals. Both categories of users thought the framework was capable and saw it as having the possibility as a useful navigation tool in future.