Algorithm 1. Object detection for visually impaired person after training of dataset.

Input: Captured image from the Camera Output: Audio for the label of the detected object Step 1:Save the captured image, I Step 2:Pre-processing of the image     Resize the image in dimensions w*h      where, w = number of pixels in the x-axis       h = number of pixels in the y-axis      Increase the contrast value for I Step 3:Load the trained deep-learning model and its parameters Step 4:Image I is processed with the deep-learning model     detections = detectObjectsFromImage(input_image = I) Step 5:Save processed output image, O       fordetections         Bounding Box prediction (bx, by, bw, bh)         Percentage probability of the object         Label, l = name of the detected object         Text to speech conversion for l       end