Minimum resolvable contrast measurement of low signal-to-noise ratio image recognition based on convolutional neural network

Yu Wang; Yanyang Liu; Peng Zhao; Xulei Qin; Ye Tao; Ye Li

doi:10.1117/12.2605315

24 November 2021 Minimum resolvable contrast measurement of low signal-to-noise ratio image recognition based on convolutional neural network

Yu Wang, Yanyang Liu, Peng Zhao, Xulei Qin, Ye Tao, Ye Li

Author Affiliations +

Proceedings Volume 12065, AOPC 2021: Optical Sensing and Imaging Technology; 1206517 (2021) https://doi.org/10.1117/12.2605315
Event: Applied Optics and Photonics China 2021, 2021, Beijing, China

Abstract

A new objective measurement method of minimum resolvable contrast (MRC) based on convolutional neural network (CNN) is proposed in this paper, in view of the fact that the subjective measurement results are easily affected by the observer’s subjectivity. Due to the low signal-to-noise ratio (SNR) of the low-light-level (LLL) images, it is difficult for traditional recognition algorithms to achieve ideal results, but the CNN can automatically learn features from the sample data for image recognition. This method does not depend on subjective judgment. It uses neural network instead of human eyes to recognize low SNR LLL images with different spatial frequencies and contrasts. The experimental results show that CNN is accurate and reliable, MRC images can be effectively recognized by it. The objective measurement of MRC based on CNN has good stability.

Citation Download Citation

Yu Wang, Yanyang Liu, Peng Zhao, Xulei Qin, Ye Tao, and Ye Li "Minimum resolvable contrast measurement of low signal-to-noise ratio image recognition based on convolutional neural network", Proc. SPIE 12065, AOPC 2021: Optical Sensing and Imaging Technology, 1206517 (24 November 2021); https://doi.org/10.1117/12.2605315

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available