Abstract

In public cloud anyone can see data of any user but in private cloud data is secured and limited to particular user. First, large-scale image recognition requires the storage of large image dataset, and the corresponding annotation contents are also huge in scale and size. Second, the image recognition task itself is computation intensive and consumes a lot of time and energy. Smart campus images can be stored in cloud servers and image features can also be extracted through cloud computing to alleviate storage and computation problems. Nevertheless, there are many challenges that need to be solved, such as the security problem of images in cloud computing in this paper, we proposed privacy-preserving image pixel selecting for secure cloud data sharing and cross domain access control is used to access the data efficiently. In this method the security system is based on selecting the accurate pixel in shown image. In cross domain access control to access the data efficiently the multi matching is performed. The façade watermarking least algorithm (FWL) used is to compress compressed bit stream to form a full resolution in the pixel domain. The proposed used to protect the data when process is while storing the data’s in cloud with the image based on pixel which selected by the transmitter.

Paper Details