基于光场成像的大景深、远距离虹膜采集方法研究

Research on Large Depth of Field and Long-Distance Iris Acquisition Method Based on Light Field Imaging

  • 摘要: 虹膜识别技术是安全性最高的生物特征识别技术之一,但受限于虹膜尺寸小、易被遮挡、常处于运动中的生理结构特点,使采集问题成为限制虹膜识别发展的最大瓶颈。尤其在远距离、复杂环境下进行虹膜识别时,系统采集容量很小、成像极易失焦,识别成功率低。鉴于此,本文将光场成像与虹膜识别技术相结合,设计基于光场成像的大景深、远距离虹膜采集方法,并与相同距离下的传统相机的虹膜采集方法进行了对比。本文设计的基于光场成像的大景深、远距离虹膜采集方法,可实现0.4~1.5m范围内,近、中、远距离虹膜实时采集。并通过光场成像的数字重对焦技术,极大提升了虹膜采集深度,近距离拍摄时( 0.4 m)将虹膜采集容量从1 cm提高至17 cm;中距离拍摄时( 0.7 m)将虹膜采集容量从18 cm提高至37 cm;远距离拍摄时( 1 m)将虹膜采集容量从10 cm提高至16 cm。本文从虹膜图像获取的源头创新,通过光场成像的景深拓展与数字重对焦技术,有效解决了传统虹膜采集设备采集容量小、难以精确对焦、需用户高度配合,导致识别成功率低的问题,为远距离、非配合式虹膜采集提供了新的思路和方法。

     

    Abstract: Iris recognition technology is one of the most secure biometric identification methods. However, due to the small size of the iris, its tendency to be occluded, and its physiological characteristic of frequent movement, data acquisition remains the greatest bottleneck restricting the development of iris recognition. In particular, during long-distance recognition under complex environments, iris images are difficult to acquire, with limited system capacity and a high risk of defocus, resulting in low recognition success rates. To address these challenges, this paper combines light field imaging with iris recognition and proposes a large depth of field and long-distance iris acquisition method based on light field imaging. Comparative experiments with conventional long-distance infrared imaging devices at the same distances were conducted. The proposed method enables real-time iris acquisition within a range of 0.4 m to 1.5 m, covering close, medium, and long distances. Furthermore, by leveraging the digital refocusing capability of light field imaging, the depth of iris acquisition is greatly enhanced: for close-range imaging (0.4 m), the acquisition capacity increased from 1 cm to 17 cm; for medium-range imaging (0.7 m), from 18 cm to 37 cm; and for long-range imaging (1.5 m / 1 m), from 10 cm to 16 cm. This work introduces innovation at the source of iris image acquisition, and through the expanded depth of field and digital refocusing enabled by light field imaging, effectively addresses the limitations of traditional systems, such as small acquisition capacity, difficulties in accurate focusing, the need for high user cooperation, resulting in low recognition success rates. The proposed approach provides a new perspective and method for long-distance, passive iris acquisition.

     

/

返回文章
返回