VIDEO SURVIVALLANCE BASED FACE RECOGNITION USING FLTP

摘要

Recognition of Face in a group of people is a bit of difficult in now days, in this paper, a new method called Fuzzy Logic Local Ternary Pattern has been introduced. This FLTP method is a commanding technique to identify the faces clearly and even their emotions too. Here, several videos are taken in to the database and compares with query image. Using FLTP, recognizes the person is present in those videos or not. It also recognizes the emotions of that person. Several techniques developed to recognize the face i.e., LBP, LTP, DRLBP, all these have certain disadvantages, hence the proposed method overcomes the problems and this FLTP is the efficient method to recognize the faces accurately and their emotions too. In the experimental results, taken five videos of different persons and compares the query image with the frames of database videos. It recognizes whether the person is present in those videos or not. Background is fixed and this algorithm is capable of identifying individual persons in a group video too. The values of FLTP on database frames and query image are taken and graphs are plot. This project can extend to security like banking sector, military, malls, we used to keep cc footages, by using this FLTP method we can easily recognize the thief or wanted person. Works are going on to recognize the face who is wearing mask.