Shape Verification Neural Network to Detection Personal Signature
Adil Alrammahi1, Hind Shaban2 

1Prof. Adil AL-Rammahi. Department of Computer Science and Mathematics, University of Kufa, Iraq.
2Prof. Hind Shaban, Department of Computer Science and Mathematics, University of Kufa, Iraq.

Manuscript received on 09 March 2019 | Revised Manuscript received on 17 March 2019 | Manuscript published on 30 July 2019 | PP: 5125-5129 | Volume-8 Issue-2, July 2019 | Retrieval Number: B2817078219/19©BEIESP | DOI: 10.35940/ijrte.B2817.078219
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: This paper was concerned to study the standard features of the personal signature image. Independent features such as the area, height, width, maximum vertical, maximum horizontal, and fractal dimension were taken to isomorphic of the original image. for more efficiently of this work the degree of complexity of image was studied and introduced as extra property of signature. Proposed algorithm was introduced for calculating the features of signature image. Many test image were taken for test this algorithm The executive appears that this algorithm is very sensitive and accurate when we have different and similar signature images. In this paper a method of verifying signatures using a technique of neural network is presented. static of various and dynamic of signature features are added in this model. Our proposal algorithm enhanced via neural network and Elman network. In our modified our method, many parameters of detection of digital images were added. These image parameters named as Width, Area, Maximum Vertical Projection, and Fractal Dimension. Fractal dimension is considered as the degree of complexity of the shape. Fractal dimension calculated the condensation of complexity of pure points of the shape without its boundaries. The dimension of the signature is varying between 1.0025 to 1.0584. Originally the signature is belonging to R2-Space (with dimension 2). The signature is considered as one of to the fractal shape. The fractal dimension is calculated by the Theorem of Boxes Squares Counting. The fractal signature dimension added an accuracy to our method for detection many signature shapes by the number of similarities in certain properties.
Keywords: Fractal Dimension, Image Processing, Neural Network, Elman Neural Network.
Scope of the Article: Image Processing and Pattern Recognition