Light Interference Fringe Detection with DBSCAN for Integrated Mirrors

This paper presents a method for detecting light interference fringes and designing integrated optical mirrors using DBSCAN clustering. Conventional edge detection methods struggle with gradual intensity changes in fringe patterns, but DBSCAN effectively identifies peaks while filtering noise. The detected fringe profiles are processed and input into FDTD simulations to analyze the interaction with an aluminum mirror. The results show 78% upward transmission efficiency and a clear far-field distribution. This approach improves mirror design for applications in communication systems and photonic circuits.