Po-Shao Chen, Yen-Lung Chen, Yu-Chi Lee, Zih-Sing Fu, Chia-Hsiang Yang
{"title":"A 28.8mW Accelerator IC for Dark Channel Prior Based Blind Image Deblurring","authors":"Po-Shao Chen, Yen-Lung Chen, Yu-Chi Lee, Zih-Sing Fu, Chia-Hsiang Yang","doi":"10.1109/A-SSCC53895.2021.9634738","DOIUrl":null,"url":null,"abstract":"Image deblurring can be applied to mitigate the effect introduced by relative movement between object and camera, as shown in Fig. 1(a). Blind deblurring restores sharp images without knowing the camera motion, but it is also very time consuming. For example, it takes 11.4 minutes to process a full-HD image for [1]. Since the blur kernel estimation occupies 99% of blind deblurring process, a dedicated accelerator that utilizes the prior knowledge of gradients for real-time deblurring on mobile devices is presented in [2], which reduces the processing time from 11.4 minutes to 1.7 seconds.","PeriodicalId":286139,"journal":{"name":"2021 IEEE Asian Solid-State Circuits Conference (A-SSCC)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2021 IEEE Asian Solid-State Circuits Conference (A-SSCC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/A-SSCC53895.2021.9634738","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Image deblurring can be applied to mitigate the effect introduced by relative movement between object and camera, as shown in Fig. 1(a). Blind deblurring restores sharp images without knowing the camera motion, but it is also very time consuming. For example, it takes 11.4 minutes to process a full-HD image for [1]. Since the blur kernel estimation occupies 99% of blind deblurring process, a dedicated accelerator that utilizes the prior knowledge of gradients for real-time deblurring on mobile devices is presented in [2], which reduces the processing time from 11.4 minutes to 1.7 seconds.