A Mobile Computing Method Using CNN and SR for Signature Authentication with Contour Damage and Light Distortion

Table 3

SIA algorithm for signature authentication.

Input:	Training samples of the true signature images, testing samples of the signature images, and the given error parameter .
Output:	The recognition result for the unknown signature.
Step 1:	Filter each sample by using the proposed golden G-L filtering algorithm and obtain the sample set .
Step 2:	Segment each sample of using the developed MPT segmentation algorithm and obtain the sample set .
Step 3:	Design the CNN structure parameters as per Figure 3.
Step 4:	Build the CNN using the sample set and obtain the true signature feature images of and from the designed CNN.
Step 5:	Calculate the over-complete dictionary and the sparse coefficient nonzero solutions N by using and .
Step 6:	Reconstruct the true signature template .
Step 7:	Reconstruct the unknown signature using the same method as the true signature template SR.
Step 8:	If , then the test signature is classified as true. Otherwise, the test signature is classified as false.