Scene Background Initialization (SBI) dataset The SBI dataset has been assembled in order to evaluate and compare the results of background initialization algorithms. It is part of the material prepared for benchmarking background initialization [1], a research work concerning the systematic comparison of some of the existing background initialization methods surveyed in [2]. It includes: A) a dataset of 7 image sequences and the corresponding 7 ground truth backgrounds; B) Matlab scripts for evaluating background initialization results. Prospective authors of the SBMI2015 Workshop, as well as any other researcher involved in background initialization, are invited to conduct and report results of quantitative evaluation of their methods using the above material. [1] L. Maddalena, A. Petrosino, "Benchmarking Background Initialization", submitted in 2015. [2] L. Maddalena, A. Petrosino, "Background Model Initialization for Static Cameras", in Thierry Bouwmans and Fatih Porikli and Benjamin Hoferlin and Antoine Vacavant (Eds), "Background Modeling and Foreground Detection for Video Surveillance", Chapman & Hall/CRC, 2014.

Related datasets