Detection of the Authenticity and Originality of Signatures in Electronic Images of Higher Education Documents Using Morphological Operations

Detection of forgery or originality of scientific documents has been one of the most exacting challenges so far. Using the prominent methods and the latest technologies and computer softwares, the forgers have been able to achieve their illegal goals. In most of the cases, the document has been forged in a way that it is difficult even for the experts to detect the original document from the forged one and at times they make a mistake. The signature and the seal inserted at the bottom of every document show its being official and authenticated by the organization that issues it. The purpose of the present study was to detect the signature in the electronic images of the documents issued by higher education institutes, such as diplomas. The study was conducted using image-processing techniques and morphological operations applied on electronic images of documents. At first, the document image was fitted and oriented on the X and Y axes. Then, through some steps, the preprocessing operations eliminated the noise and normalization operations were performed on the image. Following this, the separate parts of the signature were connected to each other using morphological operation and the signature was extracted from the image in an integrated form. Finally, a number of features were extracted from the integrated signature and the signature was detected through these features. The proposed method performed with 93% accuracy and 89% precision better than the other methods reviewed in this article.