High-speed Optical Tracking System for Dynamic Iris Authentication

Summary

The iris in biometric authentication has advantages such as non-contact measurement and low possibility of damage, but it is highly constraining to humans because it requires the user to approach and remain still at the measurement device to obtain a high-resolution image of the iris features. In order to realize remote iris authentication for humans in motion, such as walking, it is necessary to solve the trade-off between angle of view and resolution in the camera and the focus position issue. In this study, we propose a system that captures well-focused iris images by sinusoidal scanning of a liquid lens equipped with a high-resolution camera while using galvanometer mirrors to track a narrow area near the eye at high speed (Fig. 1).

We also propose a method to calibrate the response of the liquid lens and optimize the sinusoidal scanning parameters (Fig. 2). By comparing the focus of a high-speed camera equipped with the liquid lens and an imaging lens at different depth positions, such as the edge of a flat checker pattern, the response of the liquid lens is measured in milliseconds and features such as the frequency response are extracted. Then, the appropriate sinusoidal parameters are derived through iterative optimization of the scanning range, taking into account the human walking forward speed, to realize zoom adjustment and sinusoidal focus tracking that follows the movement of the target (Fig. 4). This technology is expected to be applied to the security field, such as smooth security gates.

This content is based on the results of joint research with NEC Corporation.

Fig. 1 Concept of high-speed optical iris tracking.	Fig. 2 Response calibration for scanning focus tracking.
Fig. 3 Liquid lens response measurement.	Fig. 4 Scanning focus tracking.

Reference

1. Tomohiro Sueishi, Keiko Yokoyama, Shoji Yachida, and Masatoshi Ishikawa: Image-based Response Measurement of Liquid Lens and Iterative Calibration of Scanning Focus Tracking for Dynamic Iris Authentication, 2025 IEEE/SICE International Symposium on System Integration (SII2025) (Munich, 2025.1.23)/Proceedings, pp.1016-1021
2. Tomohiro Sueishi, Arata Jingu, Shoji Yachida, Michiaki Inoue, Yuka Ogino, and Masatoshi Ishikawa: Dynamic Iris Authentication by High-Speed Gaze and Focus Control, 2021 IEEE/SICE International Symposium on System Integration (SII2021) (Fukushima, 2021.1.14)/Proceedings, pp.813-814