[8p-A23-1~12] CS.2 3.2 情報フォトニクス・画像工学、4.4 Information Photonicsのコードシェアセッション

16:15 〜 16:30英語発表

[8p-A23-8] Speckled-learned Recognition of Orbital Angular Momentum using Obstructed Speckle Patterns

Ravi kiran Jadi¹, Purnesh Singh Badavath¹, Chayanika Sharma¹, 〇(P)Vijay Kumar¹ (1.NIT Warangal India)

キーワード：

Optical angular Momentum、Speckle、Machine learning

Orbital angular momentum (OAM) modes offer a promising platform for increasing the information-carrying capacity of optical communication systems. Speckle patterns provide a compact representation of OAM information and can be directly analyzed using machine learning techniques without requiring complex field-reconstruction procedures. In this work, OAM mode recognition is performed from speckle intensity distributions using a convolutional neural network in the presence of obstacles. A grayscale image transmission experiment is further presented to demonstrate reliable information recovery through obstructed optical channels.