Image Compression and Reconstruction Based on PCA

In view of the disadvantages of large image space, high dimension of feature representation and large storage, this paper uses principal component analysis to compress the data, which can effectively reduce the loss of information, reduce the dimension of data and extract features in all aspects of image compression. In this paper, based on the analysis of PCA algorithm dimension reduction, reconstruction, feature extraction principle, the PCA algorithm is applied to image compression and reconstruction. Through the principal component analysis of sample variables, the feature vector of sample variance is calculated, and its principal component is extracted, and the contribution rate is calculated to realize image PCA compression. By extracting different eigenvalues from the original image and calculating its compression ratio and contribution rate, we can see that the clarity of the image is positively correlated with the value of the eigenvalue, and the larger the value of the eigenvalue, the higher the contribution rate. PCA based image processing technology is simple to use, high compression rate and high quality of image reconstruction, which is unmatched by many other methods.