A Secure Image-Video Retrieval Scheme with Attribute-Based Encryption and Multi-feature Fusion in Smart Grid

Abstract

AbstractWith the continuous development of smart grid, massive sensitive videos of power grid are stored in semi-trusted cloud servers, which leads to various security threats. And the existing secure video retrieval solutions are difficult to achieve both high retrieval rate and accuracy. We propose a novel secure video retrieval scheme in smart grid based on image content, by utilizing symmetric searchable encryption technology with attribute-based encryption and multi-feature fusion. Fine-grained access control is achieved for sensitive video data to eliminate illegal authorized access. This scheme converts the video into a keyframe set and transforms the difficult video secure retrieval problem to the more mature image secure retrieval region, which can effectively reduce the overhead of video secure retrieval and improve the retrieval accuracy. Security analysis demonstrates that the proposed scheme can preserve the privacy of data, resist known ciphertext attacks, known background attacks and collusion attacks. The experimental results indicate that the proposed scheme has the best balance between retrieval efficiency and retrieval accuracy among the three comparison schemes. The proposed scheme is feasible in the smart grid application scenario.KeywordsSmart gridCiphertext-policy attribute-based encryptionImage content-based video retrievalLocality-sensitive hashMulti-feature fusion