A robust automated system for detecting and recognising the digit of electrical energy consumption number of the postpaid kWh-meter

Most of the processes of kilowatt-hour meter (kWh-meter) reading in Indonesia are still in manual process which may lead to some problems, such as time consumption and high possibility of data entry errors. Therefore, this study proposes an automated system to minimise these problems. This system is developed for the image with uneven illumination condition and tilted position of stand kWh-meter due to the unavoidable situation while capturing the kWh-meter image. In this study, the illumination problem is solved by local thresholding and the tilted position of stand kWh-meter is solved by combination of morphology operations and vertical edge detection on the location detection process and vertical-horizontal projections on the segmentation process. Finally, the numeral recognition is performed by support vector machine (SVM) classifier with zonal density feature as a selected input. The results show that the accuracy of proposed system is 93.55% on detection location process, 89.38% on segmentation process, and 78.10% on numeral recognition process.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.