Content Based Image Retrieval with Enhanced Privacy in Cloud Using Apache Spark

Content Based Image Retrieval (CBIR) is the application through the integration of various traditional computer vision techniques to the image retrieval problem. CBIR usually contains high volume of personal and authenticated information which makes image privacy as a major concern. The proposed scheme for privacy allows the data owner to outsource the image database and the CBIR service to the cloud, without revealing the actual content of the database to the cloud server. The system uses only the index of the query image to prevent revealing the content of the image. Second, the image is encrypted using the session key generated between the client and the cloud server and then sent to the server for computation. Being the reduced information, it is computing stiff for the server to know the client’s interest. The proposed architecture also uses a novel framework called Apache Spark which is considered to 100x faster than MapReduce in certain applications. Result shown Apache Spark is in-memory computations which helps a lot in the retrieval of similar images. Apache Spark framework has been used when the images are newly inserted into the Cloud storage at the point of which features are extracted from the image using transformation and action functionalities and stored in the feature database.