A study on reflectance estimation for hyperspectral imaging and its application to image restoration

2017 Fiscal Year Final Research Report

• Project/Area Number: 16H07021
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Single-year Grants
• Research Field: Perceptual information processing
• Research Institution: Kagawa University
• Principal Investigator: Matsuoka Ryo 香川大学, 工学部, 助教 (40780391)
• Project Period (FY): 2016-08-26 – 2018-03-31
• Keywords: スパースコーディング / 凸最適化 / 非凸最適化 / 画像復元

Outline of Final Research Achievements

Storing scene reflectance, which independents to environmental light sources, as an image (reflectance estimation) is an important issue for the development of digital archive and medical technology. However, a reflectance estimation technique still has not been established for hyperspectral imaging having high spectral resolution. A major factor is a lack of light intensity when taking an HS image, and it causes sensor noise, focus blur, under-exposure.
To solve these problems, this research established novel multiple image blending methods based on convex optimization and sparse coding techniques.

Free Research Field

画像処理, 信号処理