Abstract

The text extraction has been developing rapidly and is an important research field in recognition from meter images. This paper brings out an algorithm to extract the text from a water meter image. The meter image has same scales of the width and height of the target area which makes this algorithm suitable to recognize the numerical characters. The algorithm finds the region of interest (ROI) of water meter using MatPlotLib, then all numeric are segmented. The segmented digit, each row value is added and stored as an array values from 0-9. To extract text from a water meter image, first segments the digits and each row value is added and compared with previous stored array values. The comparison is done by finding the Euclidean distance of two arrays and then computed the text information. The algorithm is able to extract text from the specified meter images using OpenCV- Python.