Photo sharing is an attractive feature which popularizes on line Social Networks (OSNS). Sadly, it could leak users' privateness if they're allowed to publish, remark, and tag a photograph freely. we strive to address this issue and have a look at the state of affairs while a person shares a photograph containing individuals other than him/her (termed co-image for brief).To save you viable privateness leakage of a photograph, we layout a mechanism to allow each individual in a image be aware of the posting interest and participate within the choice making at the photo posting. For this motive, we need an efficient facial recognition (FR) machine which can apprehend all of us within the picture. But, extra stressful privateness putting may additionally restriction the variety of the images publicly available to teach the FR gadget. To address this predicament, our mechanism attempts to utilize users' private pics to design a personalized FR gadget in particular trained to distinguish viable image co-proprietors without leaking their privateness.We additionally broaden allotted consensus based technique to reduce the computational complexity and guard the personal schooling set. We show that our device is superior to other viable methods in phrases of reputation ratio and performance. Our mechanism is implemented as evidence of concept Android software on Face book's platform. The energy-regulation distribution is caused by the preferential attach technique, in which the possibility of a person A connecting to a user B is proportional to the range of B's current connections.