Confused Modulo Projection based Somewhat Homomorphic Encryption -Cryptosystem, Library and Applications on Secure Smart Cities

With the development of cloud computing, storage and processing massive visual media data has gradually transferred to server. For example, if intelligent video monitoring system cannot process a large amount locally, will be uploaded cloud. Therefore, how in without exposing original become an important research topic. We propose single-server version somewhat homomorphic encryption cryptosystem based on confused modulo projection theorem named CMP-SWHE, which allows server complete blind seeing effective information user data. On client side, is encrypted by amplification, randomization, setting confusing redundancy. Operating side equivalent operating As extension, we designed implemented computing scheme accelerated batch technology improve efficiency. To make this algorithm easy use, also efficient general library CMP-SWHE. have applied foreground extraction, optical flow tracking, object detection with satisfactory results, are helpful for building smart cities. discuss extend deep learning applications. Compared other cryptosystems libraries, results show that our method obvious advantages Although some tiny errors ( $10^{-6}$ ) when too large, it very practical, especially suitable image processing.