principal component analysis VS Linear discriminant analysis template Matching

Hey folks

im doing a project with image matching. i wanna match image i meant input a image and find exact matching in database.

So
i wanna know in Image Matching what is the best techniques between
principal component analysis VS Linear discriminant analysis template
Matching and Eigen based matching also Feature-based method for matching

i
have read definition and some research paper about these algorithms but
i cant understand which is the best method in my scenario. i meant
these techniques(principal component analysis VS Linear discriminant
analysis template Matching and Eigen based matching also Feature-based
method for matching) i wanna know the exact usage (when to use )of these
algorithms.

All research papers mentioned these techniques are for matching and these techniques use every where. so im little bit confused.

So please can anyone tell me WHAT IS THE EXACT DIFFERENCE OF THESE METHOD AND WHEN TO USE and pros and cons. and justifications

thank you.