An improved performance of reversible data hiding in encrypted images using decision tree algorithm

Abstract

The advancements in increase of internet technology, transfer of data through insecure channel has become more popular. Now a day's most of the customers were turned to digital world and the channel is the main resource of communication. There is no guarantee that stored data will not be accessed by any hacker or cloud service provider. Privacy and Security is the main aspect of the rapid development in the digital storage. These conditions demand to make sure that the data is encrypted, stored and transmitted. Reversible Data Hiding in Encrypted images using machine learning is an emerging methodology used in the domain to secure the information for end-to-end secure data transmission. In this proposed method, all the data's can be Encrypted using Digital Keys, and these data can be embedded in cover images using decision tree algorithm. In the receiver side, the data can be retrieved back if receiver have right keys which is used while encryption. This system uses pseudo random number generation for generating multiple keys to encrypt and decrypt the secret data. This method is designed for encryption the Images and Text data. To increase further security and embedding capacity a machine learning algorithm is also implemented. The proposed Decision Tree method is tested on the standard image processing images and certain parameters are calculated for performance metrics and compare the results with existing schemes. The Algorithm is implemented in MATLAB and used the Statistics and Machine Learning library in MATLAB.