Small target detection method with high accuracy for visible colony RGB image formed by bacteria in water

Shousheng Liu; Zhigang Gai; Mei Zhang; Fengxiang Guo; Xu Chai; Yibao Wang; Ding Hu; Shaoyan Wang; Lili Zhang; Xueyu Zhang; Zhigang Chen; Xiaoling Sun; Xin Jiang

doi:10.1117/12.2592262

26 January 2021 Small target detection method with high accuracy for visible colony RGB image formed by bacteria in water

Shousheng Liu, Zhigang Gai, Mei Zhang, Fengxiang Guo, Xu Chai, Yibao Wang, Ding Hu, Shaoyan Wang, Lili Zhang, Xueyu Zhang, Zhigang Chen, Xiaoling Sun, Xin Jiang

Author Affiliations +

Proceedings Volume 11767, 2020 International Conference on Optoelectronic Materials and Devices; 117671D (2021) https://doi.org/10.1117/12.2592262
Event: 2020 International Conference on Optoelectronic Materials and Devices, 2020, Guangzhou, China

Abstract

The detection of microorganisms like bacteria in water is extremely important and challenging. Colony detection is an effective solution because visible colonies can be formed by bacteria in water. Aiming at the characteristics of small targets and regular shape in RGB image, a cascade network combining SSD_MOBILENET_V1_FPN with SVM HOG classifier is designed to detect and count the colonies with high accuracy, both error rate and missing rate were less than 3%. To solve the performance bottleneck, a on-chip system device Zynq UltraScale+ MPSoC EV is applied to accelerate colony count cascade network based on FPGA which could support AI computing acceleration, it can detect and count 10 colony images per second with the advantage of portability and low power consumption.

Citation Download Citation

Shousheng Liu, Zhigang Gai, Mei Zhang, Fengxiang Guo, Xu Chai, Yibao Wang, Ding Hu, Shaoyan Wang, Lili Zhang, Xueyu Zhang, Zhigang Chen, Xiaoling Sun, and Xin Jiang "Small target detection method with high accuracy for visible colony RGB image formed by bacteria in water", Proc. SPIE 11767, 2020 International Conference on Optoelectronic Materials and Devices, 117671D (26 January 2021); https://doi.org/10.1117/12.2592262

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