The resource scarcity of mobile devices, pulls them back from executing the computation intensive computer vision applications like face recognition, augmented reality, video processing etc. As a solution cyber foraging, where mobile devices offload either computation or data or both to a high performance system within their network, can be applied. Cloudlet is a new emerging technique that scales up the computational ability of mobile devices to handle the high computation requirement. It allows mobile client to communicate with nearby available server and the minimal processed data can be transferred from mobile app to the server which will do the necessary processing and result can be send back to mobile in real time. This paper explains two novel approaches of successful implementation of principal component analysis (PCA) based face recognition using VM Based Cloudlet on mobile device.