Research on the network substitution effect of Mask R-CNN algorithm for fuzzy input images

Jiang Shan

doi:10.1117/12.2671332

28 April 2023 Research on the network substitution effect of Mask R-CNN algorithm for fuzzy input images

Jiang Shan

Proceedings Volume 12610, Third International Conference on Artificial Intelligence and Computer Engineering (ICAICE 2022); 126101Y (2023) https://doi.org/10.1117/12.2671332
Event: Third International Conference on Artificial Intelligence and Computer Engineering (ICAICE 2022), 2022, Wuhan, China

Abstract

Mask R-CNN is a relatively mature method for instance segmentation at this stage, aiming at the problems of segmentation boundary accuracy and poor robustness to blurry pictures in the Mask R-CNN algorithm, an improved Mask R-CNN instance segmentation method is proposed. Use the SegNeXt network structure to optimize the mask branch for further segmentation of candidate regions, and then use new anchor size and IOU standards so that the candidate box can cover all instance regions. Finally, a method is used to add partially transformed data from a transformation network for training. Compared with the original algorithm, the total mAP value is increased by 2.2%, and the accuracy and robustness of the segmentation boundary are improved.

Citation Download Citation

Jiang Shan "Research on the network substitution effect of Mask R-CNN algorithm for fuzzy input images", Proc. SPIE 12610, Third International Conference on Artificial Intelligence and Computer Engineering (ICAICE 2022), 126101Y (28 April 2023); https://doi.org/10.1117/12.2671332

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