International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056 Volume: 02 Issue: 02 | May-2015 www.irjet.net p-ISSN: 2395-0072 © 2015, IRJET.NET- All Rights Reserved Page 583 DETECTION OF TAMPERING IN COLOR IMAGE Manoj Nagar 1 , Pinky Brahmbhatt 2 , Dr. M. Sarada Devi 3 1 PG Student, Department of ECE, L. D. College of Engineering, Ahmedabad, Gujarat, India 2 Associate Professor, Department of ECE, L. D. College of Engineering, Ahmedabad, Gujarat, India 3 Principal, Department of ECE, Narnarayan Shastri Institute of Technology, Jetalpur, Gujarat, India ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - In the last few years there is a tremendous development in the area of high quality digital camera technology. So our life is full of the use of these digital images. However now a days there are lot of software (for example Photoshop, Photoscape, Photoplus and Picasa etc.) that can be used to modify these digital image. Therefore we cannot use these images as a proof or evidence. Therefore detection of tampering in image is important issue for forensic department. In this paper I have presented a method which is based on digital watermark for detection of tampering in image. In this method I have first embed a digital watermark in Least Significant Bit of pixel which is computed from digital content of image. My proposed algorithm consists of two parts. First is generation and embedding of digital watermark and second is detection of tampering and localization. Key Words: Digital watermarking, Spatial domain, Frequency domain, Least Significant Bit (LSB), Discrete Cosine Transform (DCT), Discrete Fourier Transform (DFT). 1. INTRODUCTION Detection of tampering means showing the alteration which is not easily observable. In the last few years, malicious attackers generally try to alter meaningful information of an image so that meaning of image is changed. Watermarking methods are desirable because these methods protect the integrity of image and provide authentication. Therefore, watermarking method for detection of tampering has much attention. In Digital watermarking method additional data is added into the digital content of image in such a way that distortion caused by embedding data remains imperceptible. The additional data which is embedded is called DzDigital Watermarkdz. Digital watermarking method is used for tamper detection and recovery. Embedding reference pattern (watermark) into image [1], the problem of integrity and authentication of image is solved and attracted the interest of researchers. Many watermarking methods can determine that image has been modified or not and some methods can localize the modified areas and some methods has capability to recover tampered areas [2-6]. Generally there are three types of question arises related to detection of tampering in image [8].  Work altered in any way?  Work altered significantly?  Which parts of the Work altered? There are many methods exists for giving the answer of above question for example first question can be answered by using cryptographic signature to the content of image. Even without using cryptographic signature tampering can be detected using detection modification to the images, identifying anomalies such as variation in the shadow, missing shadow, discontinuities in background and variation of lighting condition etc. But there are some potential benefits are using watermark for tampering detection. First, watermark does not need to store as a separate, associated metadata such as cryptographic signature. Second, watermark undergoes same transformation as the content of image in which watermark is embedded. 2. LITERATURE SURVEY There are three level of tampering detection methods [9]. Low Level: This level method has no semantic information. This level method generally uses statistical property of image pixels for example DCT coefficients. Middle Level: This level method uses some semantic information. For example inconsistency of different lighting conditions and splicing etc. High Level: This level method has fully semantic information. For example image does not have semantic meaning in which toothbrush is used for hair setting. The detection of tampering in image can be done using two techniques - blind techniques and non-blind techniques. Blind tampering[15-18] detection methods are based on the fact that they do not require original image or its representation for the purpose of tampering detection whereas Non-blind [10-14] tampering detection method requires original image or its representation for the purpose of the tampering detection techniques that do not require the original image for the purpose of tampering detection. Blind tampering detection methods are also called as DzPassivedz methods and non-blind tampering detection methods are also called as DzActivedz methods.