Abstract: With the growing popularity of cloud computing, more and more users are willing to outsource their private data to the cloud. To ensure the security of data, data owners usually encrypted their private data before outsourcing them to the cloud server. Though data encryption improves the security of data, it increases the difficulty of data operating. This study focuses on the search of encrypted images in the cloud and proposes an efficient similarity retrieval scheme over encrypted images. The proposed scheme enables data owners to outsource their personal images and the content-based image retrieval service to the cloud without revealing the actual content of the image database to the cloud. The proposed scheme in this study supports the global feature based image retrieval methods under the Euclidean distance metric. Besides, rigorous security analysis and extensive experiments show that the proposed scheme is secure and efficient.