I am trying to teach myself the basics of facial recognition. I see that some resources use just distances between some points on the face (e.g., distance between 2 eyes, eyes to nose, etc). Some others use more sophisticated methods that (in my opinion) take into consideration every pixel on the face (e.g., PCA). What is the difference between these approaches. Is it enough for me to just use the distances between points on the face?
What lies at the heart of pattern recognition and pattern classification is the selection of the correct features that is used in decisions. And the most important properties of correct features are 1-ease of measurement , 2-power of discrimination. So not every feature is the same. And finding the ones which yields the best performance requires research, creativity and expertise.
In your problem of face recognition, you should first find those features which most critically depend on particularity of the member of their class, but should also be easy to measure. Geometric features provide both powerful and easy to measure examples. Therefore you can rely on them.
But depending on your aim, you may consider using other features as well, if you want a better discrimination of members. So therefore it depends on your requirements of discrimination fault tolerance.
The first approach assumes that you already have identified local features, special points on the face. This identification task is not always straightforward to perform: imagine a face with sunglasses. So it requires quite sophisticated preprocessing.
The second one uses more global features, by produce a set of optimal "fake images" (eigenfaces) built from a face database, in which each face is a specific combination of the "fake images".
Both have slightly different properties, and indeed "facial recognition" and "face classification" are not exactly the same tasks.
A useful reading could be Face recognition: eigenface, elastic matching, and neural nets
This paper is a comparative study of three recently proposed algorithms for face recognition: eigenface, autoassociation and classification neural nets, and elastic matching. After these algorithms were analyzed under a common statistical decision framework, they were evaluated experimentally on four individual data bases, each with a moderate subject size, and a combined data base with more than a hundred different subjects. Analysis and experimental results indicate that the eigenface algorithm, which is essentially a minimum distance classifier, works well when lighting variation is small. Its performance deteriorates significantly as lighting variation increases. The elastic matching algorithm, on the other hand, is insensitive to lighting, face position, and expression variations and therefore is more versatile. The performance of the autoassociation and classification nets is upper bounded by that of the eigenface but is more difficult to implement in practice.