Title:
IMAGING SYSTEM
Document Type and Number:
WIPO Patent Application WO/2021/130593
Kind Code:
A1
Abstract:
Color images from conventional imaging devices, such as image sensors, use color filters. Imaging elements are sold with color filters attached thereto, and the imaging elements are combined with lenses, etc., as appropriate, and mounted in electronic devices. Color filters reduce the amount of light reaching light receiving regions of image sensors simply by being installed so as to overlap the light receiving regions. An imaging system according to the present invention is provided with: a solid-state imaging element having no color filter; a storage device; and a learning device. From the HSV color space, the imaging system uses saturation as a criterion for selection to reduce training data, and carries out selection to achieve saturation of an appropriate distribution. Performing coloring as disclosed in the present disclosure makes it possible to perform not only coloring, but also simultaneous processing to emphasize the subject.
Inventors:
AKIMOTO KENGO (JP)
INOUE SEIKO (JP)
MISHIMA DAICHI (JP)
INOUE SEIKO (JP)
MISHIMA DAICHI (JP)
Application Number:
PCT/IB2020/061874
Publication Date:
July 01, 2021
Filing Date:
December 14, 2020
Export Citation:
Assignee:
SEMICONDUCTOR ENERGY LAB (JP)
International Classes:
G06T7/00; G06T1/00; H04N5/225; H04N5/232; H04N7/18; H04N9/07; H04N9/74
Domestic Patent References:
| WO2018197984A1 | 2018-11-01 |
Foreign References:
| JP2019128889A | 2019-08-01 | |||
| JP2019117559A | 2019-07-18 | |||
| JP2006080746A | 2006-03-23 | |||
| JP2007208481A | 2007-08-16 | |||
| JP2019118098A | 2019-07-18 |
Download PDF:
Previous Patent: SEMICONDUCTOR DEVICE AND METHOD FOR MANUFACTURING SEMICONDUCTOR DEVICE
Next Patent: NEURAL NETWORK MODEL AND LEARNING METHOD FOR SAME
Next Patent: NEURAL NETWORK MODEL AND LEARNING METHOD FOR SAME