Neural Network-based Efficient Measurement Method on Upside Down Orientation of a Digital Document

Abstract

As many digital documents are required in various environments, paper documents are digitized by scanner, fax, digital camera and specific software. In the case of a scanned document, we need to check whether the document is right sided or upside down because the orientation of the scanned document is determined by the orientation in which the paper document is placed. It is time-consuming for a person to check all the documents whether they are upside down. We propose an algorithm that can automatically determine upside down documents. The proposed artificial neural network-based method shows a high accuracy and efficiency in time for general documents. In addition, OCR-based method and CNN-based method were used to compare with the performance of the proposed method.