Privacy preserving data mining using image slicing and visual cryptography

摘要

Slicing geometry (image) into different frames and randomizing them can help in hiding a secret message (key) within the image and also preserve privacy. Authenticated users can use the relevant key and extract the original image at its destination. This field termed as visual cryptography or steganography is one of the most secure forms of message transfer available today and is highly suited for image files. The scope of this paper is to combine steganography and cryptography properties in such a way to make it harder to retrieve the image of the secret message. In this work, the image is sliced and randomized, the stored random slices upon match with a search string (query) the original image is retrieved. Multithread approach is combined in the searching process to reduce the time complexity of the search algorithm. This method is particularly advantageous to store large number of records, categorized into different groups, and each group indexed with specific keys and when used at the receiving side the particular group image(s) becomes accessible.