A robust background initialization algorithm with superpixel motion detection

作者：

Highlights：

• A low-complexity yet efficient algorithm for background initialization is proposed.

• The algorithm combines both spatial texture and temporal motion characteristics.

• An illumination change detection is adopted to alleviate illumination variance.

• The algorithm achieves top performance on the SBMnet dataset.

摘要

•A low-complexity yet efficient algorithm for background initialization is proposed.•The algorithm combines both spatial texture and temporal motion characteristics.•An illumination change detection is adopted to alleviate illumination variance.•The algorithm achieves top performance on the SBMnet dataset.

论文关键词：Background initialization,Superpixel,Motion detection,Density-based clustering

论文评审过程：Received 19 January 2018, Revised 25 April 2018, Accepted 4 July 2018, Available online 6 July 2018, Version of Record 16 November 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.07.004