2020/02/27 by Max Ruby, Ruby, Max, David S. Bolme +17 · 1 citation
Computer Science · Engineering · #Advanced Image Fusion Techniques #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #Image Enhancement Techniques #Image and Signal Denoising Methods #Image and Video Processing (eess.IV) #Machine Learning (cs.LG) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2002.12257
openalex publication_date 2020/02/27 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
Face recognition of vehicle occupants through windshields in unconstrained\nenvironments poses a number of unique challenges ranging from glare, poor\nillumination, driver pose and motion blur. In this paper, we further develop\nthe hardware and software components of a custom vehicle imaging system to\nbetter overcome these challenges. After the build out of a physical prototype\nsystem that performs High Dynamic Range (HDR) imaging, we collect a small\ndataset of through-windshield image captures of known drivers. We then\nre-formulate the classical Mertens-Kautz-Van Reeth HDR fusion algorithm as a\npre-initialized neural network, which we name the Mertens Unrolled Network\n(MU-Net), for the purpose of fine-tuning the HDR output of through-windshield\nimages. Reconstructed faces from this novel HDR method are then evaluated and\ncompared against other traditional and experimental HDR methods in a\npre-trained state-of-the-art (SOTA) facial recognition pipeline, verifying the\nefficacy of our approach.\n