The advances in image acquisition technique made recording images never easier and brings a great
convenience to our daily life. It raises at the same time the issue of privacy protection in the photographs.
One particular problem addressed in this paper is about covert photographs, which are taken secretly and
often violate the subject willingness. Covert photos are often privacy invasive and, if distributed over
Internet, can cause serious consequences. Automatic identification of such photos, therefore, serves as an
important initial step toward further privacy protection operations. The problem is, however, very
challenging due to the large semantic similarity between covert and noncovert photos, the enormous
diversity in the photographing process and environment of cover photos, and the difficulty to collect an
effective data set for the study. To overcome these challenges three contributions are used. First is to
consider a dataset of 2500 covert images and verify each image carefully. Second is to conduct user study
on how human perform to distinguish between covert and non-covert images. Third is to perform covert
photo classification algorithm that fuses various image features and visual attributes in the multiple kernel
learning framework.