Watermarking for Image Tamper Detection and Recovery Using Enhanced SPIHT Algorithm

碩士 === 國立臺灣科技大學 === 電機工程系 === 103 === Nowadays, the development of the technology of image authentication has taken an important role to verify the correctness and the integrity of image. Using the embedded watermarking to detect whether the image has been tampered, the technology of image authentic...

Full description

Bibliographic Details
Main Authors: Kuan-Liang Chen, 陳冠良
Other Authors: Nai-Jian Wang
Format: Others
Language:zh-TW
Published: 2015
Online Access:http://ndltd.ncl.edu.tw/handle/09449053286796272808
Description
Summary:碩士 === 國立臺灣科技大學 === 電機工程系 === 103 === Nowadays, the development of the technology of image authentication has taken an important role to verify the correctness and the integrity of image. Using the embedded watermarking to detect whether the image has been tampered, the technology of image authentication is capable of recovering the tampered image so as to provide the better protection for the image. The paper refers to using SPIHT (Set Partitioning in Hierarchical Trees) to generate watermarking. The more the watermarking is generated, the better the effect of tamper detection and image recovery will be. However, as the more watermarking is produced, the quality of watermarked image becomes worse. Therefore, this paper aims to enhance the ability of tamper detection and image recovery in the condition of the limited number of watermarking. Finally, this paper experimentally proves that the watermarked image proposed herein complies with the imperceptibility of watermark and the quality of visual image. The SPIHT decoded image can improve the PSNR performance by 2 ~ 3 dB on average. And, the method of tamper detection can precisely position those tampered blocks.