Reading values in electrical meter using image processing techniques

Abstract

This paper proposes the method for extracting the series numbers in the electrical meter. Initially the portion of the meter where the reading is actually located is captured by placing the camera in a fixed position in front of the meter. The image of the meter is captured in certain series of time interval. The input image converted to grayscale image and it is enhanced using image enhancement techniques. The image is then converted to the binary image by removing the connected components by using Adaptive Thresholding (AT), for the grayscale image. Then the Morphological operation on binary image is performed as to produce an enhanced binary image. The image is scanned horizontally until the white pixel is encountered and the resultant image is stored in an array matrix. Then the resultant image is segmented by Vertical Edge Detection Algorithm and each segmented image is stored in an individual array matrix. Each image is then compared with the actual template and the result is stored in the text file. It compared with the Vertical Edge Detection Algorithm to the Sobel operator in terms of accuracy, algorithm complexity, and processing time. The results show accurate edge detection performance and faster processing than Sobel by five to nine times.