| Autor: |
Lingjun Li, Liusheng Huang, Xinxin Zhao, Zhili Chen, Wei Yang, Zhenshan Yu |
| Rok vydání: |
2009 |
| Předmět: |
|
| Zdroj: |
WKDD |
| DOI: |
10.1109/wkdd.2009.105 |
| Popis: |
A new steganalysis method is proposed to detect the exists of hiddeninformation using character substitution in texts. This is done byutilizing Support Vector Machine (SVM) as a classifier to classifythe characteristic vector input into SVM. The most important step ofthis detection algorithm is the construction of a propercharacteristic vector. Under the prerequisite that the secret bitsto be embedded are uniformly distributed, the distribution of thecharacters used for hiding data has altered after steganographicprocess, thus the ratio of abnormal characters to normal charactersis different in cover texts and stego texts. Experimental resultsdemonstrate that this method can reliably detect whether there ishidden information in texts. The detection accuracy in experimentreaches as high as 96\% while the embedding data rate is merely20\%. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
