System for Reading the Product Labels to Help Blind Persons

Abstract: The paper proposes system gives a strategy to help blind persons by reading the product labels. In this system, one camera is used for capturing the product images. Blind person will only have to show product label in front of the camera, then the camera will capture the image. Raspberry pi 2 is used as a processor to process the image. System extracts the text from image by using SIFT (scale invariant feature transform) features of the image. KNN (k nearest neighbor) algorithm is used for calculating the Euclidean distance. The extracted text present at the output of the processor.

Keywords: KNN (k nearest neighbor), raspberry pi 2, SIFT (scale invariant feature transform).