Related Experiment Video
Updated: Aug 9, 2025
09:32
Development of New Methods for Quantifying Fish Density Using Underwater Stereo-video Tools
Published on: November 20, 2017
9.3K
Identification and counting of fish targets using adaptive resolution imaging sonar
Wei Shen1,2, Zhanfei Peng1,2, Jin Zhang1
1College of Marine Science, Shanghai Ocean University, Shanghai, China.
Journal of Fish Biology
|February 17, 2023
Summary
Researchers developed an automated program to track and count fish using adaptive resolution imaging sonar (ARIS). This method improves the efficiency of processing large sonar datasets for marine biology assessments.
Area of Science:
- Marine Biology
- Underwater Imaging
- Computational Ecology
Background:
- Accurate fish identification and counting are vital for marine resource monitoring.
- High-frequency adaptive resolution imaging sonar (ARIS) provides close-up underwater video, even in turbid conditions.
- Processing the large data volumes from ARIS presents a significant challenge.
Purpose of the Study:
- To develop an automated image-processing program for tracking and counting free-swimming fish.
- To address the challenge of massive data output from imaging sonars.
- To enhance the efficiency of marine biological resource assessment.
Main Methods:
- Developed an automatic program fusing K-nearest neighbour background subtraction with DeepSort target tracking.
- Utilized high-frequency adaptive resolution imaging sonar (ARIS) for data acquisition.
- Evaluated the program using four test datasets with varying target sizes, observation ranges, and sonar deployments.
Main Results:
- The automated program successfully tracked and counted free-swimming fish.
- Achieved an accuracy index of 73% and a completeness index of 70%.
- Demonstrated potential for efficient processing of imaging sonar data.
Conclusions:
- The developed automated approach can replace time-consuming semi-automatic methods for fish counting.
- Offers improved efficiency for imaging sonar data processing.
- Provides technical support for future real-time data processing in marine biology.
