Reading the test images from the dataset and using the techniques for PCA, trying to obtain the eigenvector and eigenvalues of the image in order to recognize a new image of the same type. There are various datasets available for the testing and training purposes. Here we find the Mean of all image and generate a mean and stdev using that image. Secondly, find covariance matrices and use that to find eigenvalues and eigenvectors and eventually obtain eigenfaces. Then the eigenvectors and eigenvalues are normalized. Using the eigenvectors reconstruct the image that will be termed as recognized image.