Micro-expression recognition based on local binary patterns from three orthogonal planes and nearest neighbor method

Micro-expression is a very short and rapid involuntary facial expression, which reveals suppressed affect. Recognizing micro-expression can help to accurately grasp the real feelings of people, a result that can have an important practical impact. But the scholars´ studies have demonstrated that real micro-expression is difficult to identify. There are two main restrictive factors, one is the need of a comprehensive and typical database, and the other one is the need of a suitable method. This paper proposes a combination of local binary patterns from three orthogonal planes (LBP-TOP) and the nearest neighbor method, and does experiments on a large database. At last, a reasonable result is obtained.