2D-3D face recognition using shapes of facial curves based on modified CCA method

The paper deals with the example of the 2D-3D face recognition method using the CCA approach based on detected facial curves from 3D mesh surface. The paper presents the principle of the detection facial curves from 3D range images. We have introduced a powerful technique for 2D-3D face recognition, which have been evaluated on the TEXAS face database. This proposed 2D-3D face matching method is using the CCA to learn mapping between the 3D range image and 2D-3D face images using the facial curves. This method makes it possible to match the 3D face image with enrolled 2D face image. The 2D-3D face recognition results for individual cases and different number of subjects in the test set are shown in the corresponding table. Test 2D-3D face match results gave a recognition rate for modified CCA method for dataset of 115 subjects at least 98%.