Multi-spectral pedestrian detection based on accumulated object proposal with fully convolutional networks

• Published in: 2016 23rd International Conference on Pattern Recognition (ICPR) pp. 621 - 626
• Main Authors: Hangil Choi, Seungryong Kim, Kihong Park, Kwanghoon Sohn
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2016
• Subjects: Color; Feature extraction; Finite impulse response filters; Image color analysis; Proposals; Robustness; Support vector machines

This paper presents a method for detecting a pedestrian by leveraging multi-spectral image pairs. Our approach is based on the observation that a multi-spectral image, especially far-infrared (FIR) image, enables us to overcome inherent limitations for pedestrian detection under challenging circumst...