Abstract

We propose a new method for recognizing color image sets using quaternionicGrassmannians, which use the power of quaternions to capture color informationand represent each color image set as a point on the quaternionic Grassmannian.We provide a direct formula to calculate the shortest distance between twopoints on the quaternionic Grassmannian, and use this distance to build a newclassification framework. Experiments on the ETH-80 benchmark dataset and andthe Highway Traffic video dataset show that our method achieves goodrecognition results. We also discuss some limitations in stability and suggestways the method can be improved in the future.

 

Quick Read (beta)