Encrypted JPEG image retrieval using block-wise feature comparison

• An effective retrieval scheme for encrypted JPEG images is proposed.
• The JPEG format compliance and file size preservation can be ensured.
• The feature extraction/encryption is needless on the user side.
• A simple and effective block-based feature descriptor is designed.

This paper proposes a novel scheme for encrypted JPEG image retrieval, which includes image encryption and retrieval phases. Using the scheme, the content owner encrypts JPEG images by jointly applying permutation cipher and stream cipher to their corresponding bit-streams, and then transmits encrypted versions to a database server. With an encrypted query image, although the server learns nothing about the plaintext content, it may extract local statistical feature of intra-block AC coefficients using a new feature descriptor. Subsequently, exploiting block-wise feature comparison, the server can measure the similarity between encrypted query image and database image. After that, the encrypted images with plaintext content similar to the query image are returned to the authorized user. Experimental results show that the proposed scheme can ensure both format compliance and file size preservation while providing effective retrieval service in encrypted domain.