Abstract
A tensor feature transformation and enhancement algorithm are proposed in this paper based on PCA algorithm to address the challenges in face detection from low-resolution images. A 2D-PCA face super-resolution enhancement algorithm of a second-order tensor replaces the traditional 1D-PCA method for feature transformation to improve the accuracy of feature tensor solution and reduce the computational complexity. Furthermore, a coupling PCA method is applied for resolution enhancement of the combination of global information and local information of face image. Experimental results verify the stable and superior performance of the proposed algorithm.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
