Double Layer Image Security System using Encryption and Steganography

TL;DR: A hybrid image security framework has been proposed to overcome the problem stated earlier, which will be implemented by combining various techniques together to achieve the image security goal.

Abstract: The image security on internet transfers is the concern of the hour as the breaching attacks into the image databases are rising every year. The hackers take advantage of the stolen personal and important images to fulfill their dangerous and unethical intentions. The image data theft can be used to defame a person on the internet by posting the illegal and unacceptable images of that person (internet user). Hence the images transfers have to be secure to ensure the privacy of the user’s image data. In this research, a number of image security systems have been studied to evaluate the research gap. Majority of the existing image security systems are not up to date to protect against the latest breaching attacks. So, we have proposed an effective and robust image security framework particularly designed for the images. The proposed has been designed and implemented using MATLAB. In this research, a hybrid image security framework has been proposed to overcome the problem stated earlier, which will be implemented by combining various techniques together to achieve the image security goal. The techniques included in the combination would beimage compression, cryptography and steganography. DWT compression has been used, because it is a stronger compression algorithm. The steganographed image would be compressed to reduce its size. Blowfish encryption algorithm would be used for the encryption purposes. It offers maximum throughput (faster) and also energy efficient. Compressed image would be encrypted to enhance the image security. Real image will be hidden into another image. A cluster based steganographic technique will be used. Real image and face image would be analyzed, and the real image would be embedded in those areas of face image, where color schemes of the real image and face image would be most similar. Kmeans or Hierarchical clustering would be used as a clustering technique. An all new comparative analysis technique would be applied to make the comparison between real image and base image on the basis of color patterns.