Invention Grant
- Patent Title: Systems and methods for multispectral image demosaicking using deep panchromatic image guided residual interpolation
-
Application No.: US16960304Application Date: 2019-07-05
-
Publication No.: US11410273B2Publication Date: 2022-08-09
- Inventor: Zhihong Pan , Baopu Li , Yingze Bao , Hsuchun Cheng
- Applicant: Baidu USA, LLC , Baidu.com Times Technology (Beijing) Co., Ltd.
- Applicant Address: US CA Sunnyvale; CN Beijing
- Assignee: Baidu USA, LLC,Baidu.com Times Technology (Beijing) Co., Ltd.
- Current Assignee: Baidu USA, LLC,Baidu.com Times Technology (Beijing) Co., Ltd.
- Current Assignee Address: US CA Sunnyvale; CN Beijing
- Agency: North Weber & Baugh LLP
- International Application: PCT/CN2019/094839 WO 20190705
- International Announcement: WO2021/003594 WO 20210114
- Main IPC: G06T3/40
- IPC: G06T3/40 ; G06N3/08 ; G06T5/00

Abstract:
Described herein are systems and embodiments for multispectral image demosaicking using deep panchromatic image guided residual interpolation. Embodiments of a ResNet-based deep learning model are disclosed to reconstruct the full-resolution panchromatic image from multispectral filter array (MSFA) mosaic image. In one or more embodiments, the reconstructed deep panchromatic image (DPI) is deployed as the guide to recover the full-resolution multispectral image using a two-pass guided residual interpolation methodology. Experiment results demonstrate that the disclosed method embodiments outperform some state-of-the-art conventional and deep learning demosaicking methods both qualitatively and quantitatively.
Public/Granted literature
Information query