International Journal of Latest Engineering and Management Research (IJLEMR) ISSN: 2455-4847 www.ijlemr.com || Volume 04 - Issue 07 || July 2019 || PP. 61-68 www.ijlemr.com 61 | Page Unification of 2-FIS Values for Digital Image Watermarking Based on QRD Raghad I. Sabri 1 , Areej M. Abduldaim 2 1 (Department of Applied Sciences/ University of Technology, Baghdad, Iraq) 2 (Department of Applied Sciences/ University of Technology, Baghdad, Iraq ) Abstract: Fuzzy theory has become the cornerstone of many important sciences today. With the gradual transition of science into the digital world, fuzzy logic has become one of the most important influences in the development of the rest of science, including image processing. Digital watermarking techniques depends on many factors to insert the information of the watermark into the cover image. Fuzzy Inference System (FIS) is one of these factors that used to control the decision of how to embed bits of the watermark depending on the goal to be achieved. In this paper, the Mamdani type of FIS is introduced in two parallel cases utilizing two deferent inputs named contrast sensitivity and edge sensitivity respectively to achieve high robustness and acceptable imperceptibility relying on appropriate places obtained by the algebraic matrix decomposition method QR. Experimental results show that the proposed method has a promising behaver to increase the normalized correlation (NC) and the peak signal to noise ratio (PSNR). Keywords: Fuzzy Inference System; QR decomposition; watermarked image; the Peak Signal to Noise Ratio(PSNR); normalized correlation(NC). I. INTRODUCTION Fuzzy set theory and fuzzy logic are related to fuzzy mathematics which forms a branch of mathematics. Fuzzy logic is depending on the perception that people make decisions relying upon inaccurate and not numerical data, the fuzzy system is mathematical wherewithal of symbolizing obscurity and inaccurate data. These systems have the ability to representing, interpreting, recognizing, manipulating, and using information and data that are obscure and need sureness. Fuzzy theory has become the cornerstone of many important sciences today. With the gradual transition of science into the digital world, fuzzy logic has become one of the most important influences in the development of the rest of science, including image processing. Digital watermarking techniques depends on many factors to insert the information of the watermark into the cover image. Fuzzy Inference System (FIS) is one of these factors that used to control the decision of how to embed bits of the watermark depending on the goal to be achieved [1]. On the other hand, linear algebra is a subfield of mathematics interested with matrices, vectors, and linear transforms. It is a fundamental key to the field of image processing, from symbols used to describe the approach of algorithms to the enforcement of algorithms in code. Furthermore, linear algebra plays an important role in image processing, particularly in watermarking. Digital image watermarking is information (the watermark) hiding into the digital data. In other words, to affirm the originality of the data; the embedded secret image can be specified or extracted later. Digital watermarking is the first kind of mechanisms to better the impartiality and reliability of digital data. Lately, authentication is one of the major watermarking requirements in image processing applications [2]. Imran and Harvey proposed in [3] a blind adaptive color image watermarking technique depending on PCA, SVD, and HVS. To improve the perceptual quality of the watermarked image PCA is used to decorrelate the three color channels of the cover image. While the HVS and FIS worked to further improve both robustness and imperceptibility by choosing a suitable running scale, for this reason, regions more susceptible to noise can be added with additional information as compared to fewer susceptible regions. Typically the goodness of the watermarked image is handled in [4] by locating the adaptable running factor for every demarcation pixel intensity. The HVS (texture masking) and FIS were used in order to set the adaptable scaling factor. To enhance the security grade and robustness DWT has been used. This improvement is owing to the irregular apportionment of the watermark within the image through the transform converse. The algorithm of using (SVD) in order to decompose LH; and HL sub-bands is given. A novel robust watermarking scheme is implemented relying on DWT and SVD using Fuzzy Logic and Genetic Algorithm. Fuzzy logic system is used to find the strength of watermark that has to be added to the original image while embedding [5]. The essential difficulty for creating a new watermarking scheme is typically the stalemate between impressionability plus robustness. Lalani and Doye [6] proposed a technique tries to solve this problem by designing a fuzzy inference system (FIS) based on just-noticeable distortion (JND) that takes into consideration