研究目的
To verify that only one individual is in the designated transit area of a mantrap portal using a single camera system combining optical flow and machine-learning classification.
研究成果
The novel approach combining optical flow and machine-learning classification provides competitive results and outperforms detection rates in several attack scenarios, demonstrating the relevance of motion and micro movements in such a use-case.
研究不足
The system's performance may be affected by the objects carried by authorized subjects and the specific attack scenarios.
1：Experimental Design and Method Selection:
The approach combines optical flow and machine-learning classification to detect attacks in a mantrap portal.
2：Sample Selection and Data Sources:
A database was created with images of attempted attacks and regular verification.
3：List of Experimental Equipment and Materials:
A monocular camera manufactured by Point Grey with CMOS sensor was used.
4：Experimental Procedures and Operational Workflow:
Data acquisition involved a broad range of participants with different physical characteristics.
5：Data Analysis Methods:
The performance was evaluated using empirical testing and biometric evaluation metrics.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
