A Human Visual System Based Reversible Data Embedding Method for Digital Contents

View Statistics Email Alert RSS Feed

Information

Details

Project title

Code/計畫編號

NSC101-2622-E412-003-CC3

Translated Name/計畫中文名

基於人類視覺之可逆式數位媒體資訊嵌入技術

Funding Organization/主管機關

National Science and Technology Council

Co-Investigator(s)/共同執行人

洪維恩(計畫主持人)
鍾玉科

Department/Unit

Yu Da University of Science and Technology, Department and Graduate Institute of Information Management

Website

https://www.grb.gov.tw/search/planDetail?id=2596644

Year

2012

Start date/計畫起

01-06-2012

Expected Completion/計畫迄

31-05-2013

Co-Investigator(s)

Ming-Feng Yang

Bugetid/研究經費

260千元

ResearchField/研究領域

資訊科學--軟體

在本產學案中，我們提出了一個基於human visual system (HVS) 的可逆式資訊隱藏技術。Jung et al. 在2011年所提出的方法是利用前一個像素值來預測目前的像素值，並以走訪過的鄰近像素來估算目前所在的像素值其最大容許誤差，然後以控制embedding level的方式將資訊嵌入。然而Jung et al. 的方法並沒有充份利用鄰近的像素值來預測，同時它們所採用的演算法與location map的處理方法，很容易導致載量與影像品質的下降。因此，我們的方法將利用鄰近的三個像素值，來取得像素之間可容許的最大誤差，同時依照所得到的誤差來對像素值進行adaptive pre-shifting，以減少location map的大小。實驗結果可顯示，我們的方法因為利用較準確的預測法來嵌入資訊，使得影像品質有所提升，同時也可以因為location map縮小而增加嵌入的訊息。本產學案將與麗文文化事業股份有限公司合作，並將本案所開發的嵌入技術，應用在麗文所發行的數位出版品中，以確保其數位出版品的版權。 In this project, we will propose a reversible data hiding method based on human visual system (HVS). Jung et al. in 2011 proposed a reversible method that uses the neighboring pixel value to predict the current pixel values, and use the neighboring pixels to estimate the just noticeable difference (JND) of the current pixel. Data are then embedded by controlling the corresponding embedding level such that the changes between the cover and stego pixels are confined within the JND. However, Jung et al.’s method does not fully exploit the neighboring pixels for prediction and for pixel pre-shifting. As a result, the payload is decreased because the size of the location map is increased. The proposed method fully exploits the neighboring pixels and uses an adaptive pre-shifting technique to reduce the size of the location map significantly. The preliminary experimental results reveal that the proposed method will effectively increase the payload of the cover image due to a significant reduction of location map. This project will be corporate with the Li-Wen publishing corporation, and the proposed method can be used in the digital publications such as e-books and teaching supplementary for copyright protection.

Keyword(s)

人類視覺系統
直方圖位移技術
Location Map
Human Visual System
Histogram Shifting

DSpace CRIS

A Human Visual System Based Reversible Data Embedding Method for Digital Contents

Details

Description