What is the working technique of LBP for face recognition?
The idea behind using LBP features is that the face images can be seen as a composition of micro patterns which are invariant with respect to monotonic gray scale transformations. Combining these micro-patterns, a global description of the face image is obtained.
It seems you have already answered the question. LBP captures the local texture information and is robust as regards gray-scale transformation. The histogram representation also makes LBP robust as regards certain variations, e.g. translation, scaling, and occlusion.
To me, the mysterious parts of LBP include: 1) its high discriminative power in distinguishing faces of different persons; 2) why do the other texture methods, e.g. Gabor and HOG, underperform LBP? 3) why do people keep claiming that LBP is robust as regards viewpoint change and expression variation, despite the fact that LBP is theoretically variant to both anisotropic transformations and deformanational transformations.
actually this answer i have taken from the famous paper presented at ECCV 2004 by timo ahonen . the name of the paper is face recogntion with local binary patterns
LBP and its variants capture the local texture information well. Hence it is widely used in Face recognition as texture is an inherent feature that can be used to distinguish between faces. There are several variations in LBP each one is unique in what kind of texture they represent.
- Badges
- Science topic