Private Content-based Image Query System using Statistical Properties

The objective of this research work is to investigate whether content-based image retrieval system is able to respond accurately without knowing query. While using statistical properties for retrieving our query from the large database, we simultaneously try to satisfy the security problems of users. The contribution of content-based image query system to reduce human labour and time add great value to it. Privacy threats also arise in using content-based image query system because the sample of the query image is required to search the similar image from the database. This research proposes the encryption scheme that can be applied to the query images in order to preserve the privacy of the image content. Here we propose a framework for private content-based image query system that computes the statistical similarity between the image of interest and encrypted image.