A NOVEL LOW-COST IRIS RECOGNITION SYSTEM

Among biometric authentication methods, iris recognition has always received special attention from researchers over the world besides face and fingerprint recognition. However, most of the available works still lack practicality due to various reasons, mainly caused by the complexity of the methods themselves, the overall cost, andnot considering large scale deployment. In this research, we aim to build an iris recognition system consisting of multiple devices working remotely with reasonable prices and ready-to-use. Raspberry Pi 3 Model B+ is utilized as the core of hardware components. Following that idea, we designed an optimal image processing algorithm in OpenCV/C++ for the platform, and Python is chosen for application development in this project. Communication in the system’s network is supported by lightweight messaging protocol MQTT. Experiments show satisfying results in both terms of accuracy and execution time, which stimulates us to keep improving the performance of the prototype in the future.