Combining weighted curvelet accumulation with motion vector duty cycle for nonuniform video deblurring

作者：

Highlights：

• A weighted curvelet accumulation method is proposed which synthesizes the multiple adjacent frames for estimating the initial latent sharp frame.

• A motion vector duty cycle estimation method is proposed by utilizing the intro-frame correlation information and the estimated initial latent sharp frame to improve the blur kernel accuracy.

• A novel nonblind video deblurring model is built by fully utilizing the spatiotemporal information for obtaining the deblurred frame.

摘要

•A weighted curvelet accumulation method is proposed which synthesizes the multiple adjacent frames for estimating the initial latent sharp frame.•A motion vector duty cycle estimation method is proposed by utilizing the intro-frame correlation information and the estimated initial latent sharp frame to improve the blur kernel accuracy.•A novel nonblind video deblurring model is built by fully utilizing the spatiotemporal information for obtaining the deblurred frame.

论文关键词：Weighted curvelet accumulation,Motion vector duty cycle,Nonuniform video deblurring

论文评审过程：Received 30 March 2018, Revised 3 August 2018, Accepted 3 August 2018, Available online 14 September 2018, Version of Record 1 October 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.08.002