An Image Authentication Scheme Using Merkle Tree Mechanism

• Main Authors: CHEN, YI-CHENG, 陳奕呈
• Other Authors: CHOU, YUNG-CHEN
• Format: Others
• Language: en_US
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/pfkckr

碩士 === 亞洲大學 === 資訊工程學系 === 107 === Research in the digital image processing field has received significant attention and has advanced rapidly in recent years. Image tampering and misattribution have become real concerns in the open environment of the worldwide web, and scholars have proposed various image verification mechanisms to detect and mitigate image tampering. Likewise, blockchain technology has become very popular in recent years. This study proposes a novel image verification mechanism based on the Merkle tree, a fundamental component of blockchains that underpin their functionality. The Merkle tree root in the blockchain mechanism provides a reliable environment for storage of image features. The verification of images to detect tampering can be performed by the Merkle tree mechanism to obtain the hash values of the Merkle tree nodes. In addition, the proposed method combined with the Inter-Planetary File System (IPFS) to improve the availability of images. The primary purpose of this study is to achieve the goal of image integrity verification. The proposed method can not only verify the integrity of an image, but can also repair the tampered area if the image has been altered. Because the proposed method employs the blockchain mechanism, third party is not needed for image verification. The verification method is performed by each node in the blockchain network. The experimental results demonstrate that the proposed method successfully achieves the goal of image authentication and tampered area restoration. Since the verification mechanism uses the hash values for change detection, it can recognize the slightest alterations in the image. However, the tampering of the rotation and translation is less obvious.