.png){kind=icon}
.png){kind=icon}
Return
An iris quality evaluation method with pre-recognition screening function
.png){kind=icon}
PRE
AI
DOI：10.1007/s11042-021-11377-y.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
An iris image quality evaluation method for iris multi-category recognition scenes is designed in this article. The problem that the poor applicability of uniformly quality indexes and the iris physiological morphology in different states leads to poor expression of qualified image features can be solved. The method can also improve the overall effect of quality evaluation on the iris recognition process. In this method, a person with authoritative knowledge sets the qualified logic standard for qualified eye images. Every single person need to take homology to digitally express the qualified logic standard as qualified index data in each person. People are used as the evaluation unit to make the qualified indicators closer to everyone's physiological form through the category homology mode. It can avoid the difficulty of adding new person to the quality assessment and the inflexible adjustment of changes in the collection status due to the mechanization of the qualified indexes setting. The applicability of the qualified indexes can be improved. In addition, the category homology mode carry out category screening based on physiological morphology before feature expression, which can narrow the scope of iris recognition, and give quality evaluation a pre-recognition screening function. The experiment results of different iris libraries show that the qualified indexes of the method are reasonably. It can eliminate some unqualified iris categories. It can also prove that the idea of pre-screening for iris recognition in the quality evaluation process is feasible.
Keywords:
Iris image quality evaluation
Category homology
Qualified logic standard
Qualified index data
Pre-recognition screening
Journal
IF:
3
Papers:
1.9W
Citations:
3.2W
