Illegal Photography Detection with High-attack Registance Using Modulated Illumination and Single Image

2021 Fiscal Year Final Research Report

• Project/Area Number: 20K21781
• Research Category: Grant-in-Aid for Challenging Research (Exploratory)
• Allocation Type: Multi-year Fund
• Review Section: Medium-sized Section 60:Information science, computer engineering, and related fields
• Research Institution: Hokkaido University
• Principal Investigator: SUGIMOTO MASANORI 北海道大学, 情報科学研究院, 教授 (90280560)
• Project Period (FY): 2020-07-30 – 2022-03-31
• Keywords: 違法撮影判定

Outline of Final Research Achievements

The outcomes of the project are summarize as follows: (1) we find that a exposure time of a rolling-shutter camera works as a frequency filter and propose an illegal photography detection method that analyzes frequency spectra of captured images illuminated by OFDM optical signals. (2) we propose a model to explain how phases of captured images are affected by photographed objects and confirmed the model through computer simulations. (3) we conduct evaluations of the proposed method using a large-scale image database and multiple classifier algorithms, and confirm that the method can achieve the f1-score higher than 99.0 %.

Free Research Field

情報科学

Academic Significance and Societal Importance of the Research Achievements

スマートフォンの普及により、写真、動画のインターネット上での共有が容易になった。一方、それらには撮影が制限されるはずの映画館、美術館、コンサート会場等でのコンテンツも含まれ、知的財産権保護の観点から喫緊の対策が求められる。
本申請で提案、構築された違法撮影判定手法は、上記の社会的要請に応じた研究であるとともに、1枚の画像のみで実現可能である点、露光時間の周波数フィルタ機能に着目する点など、学術的にも新規性の高い点が意義として挙げられる。