Using image processing technology to create a novel fry counting algorithm

• Published in: Aquaculture and fisheries Vol. 7; no. 4; pp. 441 - 449
• Main Authors: Zhang, Jianfei, Pang, Haitong, Cai, Weiming, Yan, Zhonghong
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.07.2022; KeAi Communications Co., Ltd
• Subjects: algorithms; aquaculture; Connected area algorithm; Fish counting; fish fry; Image recognition; Thinning algorithm; transportation; Image recognition; Thinning algorithm; Connected area algorithm; Fish counting
• ISSN: 2468-550X; 2096-1758; 2468-550X
• DOI: 10.1016/j.aaf.2020.11.004

Quantitative counting of fry is required for multiple reasons, including breeding, transportation, and sales and creating a fish-counting algorithm using image processing technology is proposed as a useful method. The images for fish fry were preprocessed to extract the details of the target fish, using methods including binarization, dilation, and erosion. The thinning and the connected area algorithms were then independently adopted to count the fish fry within the image. The advantages and the disadvantages of the two algorithms under different densities were also compared, revealing that the fish fry pictures with a clear contrast between the foreground and background resulted in a higher accuracy. The main purpose of this research is to solve the problem of fry counting in the case of high density and high overlap. This research provides a reference point for this method based on image technology.