Application of Artificial Intelligence to Identify Electric Meter Tariffs

Scanning images and converting scanned information into digital format attracts significant attention from academia and industry. Scanning is an automated, fast, and efficient process compared to traditional data entry. Recognizing digits from images is a challenging task. Traditional approaches to solving this problem typically separate the localization, segmentation, and recognition steps. This paper presents a unified approach that integrates these three steps via the use of a deep convolutional neural network that operates directly on the image pixels. The proposed model is applied to the recognition of electric meter tariffs. Large dataset is used for testing and results are discussed.