This paper has enforced a spectacle model to assist visually impaired individuals with safe and economical walking within the surroundings. The walking guide uses 3 items of ultrasonic sensors to spot the obstacle in each direction, together with front, left, and right. additionally, the system will observe potholes on the paved surface victimization Associate in Nursing ultrasonic sensing element and convolutional neural network (CNN). The CNN runs on an Associate in Nursing embedded controller to spot obstacles on the surface of the road. pictures are to be trained at the start by employing a CNN on a laptop and square measure then classified on the embedded controller in a period of time. The experimental analysis reveals that the planned system has 98.2% accuracy for the front sensing element with a mistake(error) rate of 1.8% once the obstacle is at a 50 cm distance. additionally, the proposed system obtains the accuracy and loss severally for image classification. The experimental study additionally demonstrates that the developed device outperforms outstanding existing works.
Article Details
Unique Paper ID: 155272
Publication Volume & Issue: Volume 9, Issue 1
Page(s): 408 - 416
Article Preview & Download
Share This Article
Join our RMS
Conference Alert
NCSEM 2024
National Conference on Sustainable Engineering and Management - 2024