This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
No, when using it for the training stage, PCA does not require to give the same number of images for each person. Thus, you can use the different numbers of training images associated with training objects. And if you use the same number of images and different number of images for each person, the recognition rate changes because each image of one person contributes different amount of discriminative features to construct the eigenvectors. But one thing you can be sure is that: the more images you use for training, the better recognition rate you will achieve.
