Scalable face image compression based on Principal Component Analysis and arithmetic Coding
You-Ran Liu, Lih‐Jen Kau · 2017
In this paper we propose a scalable face image compression algorithm based on Principal Component Analysis (PCA) and Entropy Coding. By using PCA and some training face image patterns, we can extract the most representative eigen-image of human faces. To reduce the coding complexity as well as to achieve a higher compression ratio, only the first term of the extracted eigen-images will be used for the encoding of the human face, i.e., only the eigen-image with maximal energy strength will be selected for the encoding process. As we will see in the experiment that a good trade off between the computation complexity, compression ratio, and image quality can be achieved with the proposed algorithm.