A pond-surface-based Biofloc-farm health-monitoring system for African catfish using deep learning methods
Biofloc fish farming is becoming a popular aquaculture tool for fish farmers to farm for fishes as nutrients required for the fishes to feed on are produced naturally within the self-contained ecosystem especially in locations where naturally occurring body of water is difficult to come by and land...
Saved in:
| Main Author: | |
|---|---|
| Format: | Academic Exercise |
| Language: | en en |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/33216/1/A%20POND-SURFACE-BASED%20BIOFLOC-FARM%20HEALTH-MONITORING%20SYSTEM%20FOR%20AFRICAN%20CATFISH%20USING%20DEEP%20LEARNING%20METHODS.24pages.pdf https://eprints.ums.edu.my/id/eprint/33216/2/A%20POND-SURFACE-BASED%20BIOFLOC-FARM%20HEALTH-MONITORING%20SYSTEM%20FOR%20AFRICAN%20CATFISH%20USING%20DEEP%20LEARNING%20METHODS.pdf https://eprints.ums.edu.my/id/eprint/33216/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Biofloc fish farming is becoming a popular aquaculture tool for fish farmers to farm for fishes as nutrients required for the fishes to feed on are produced naturally within the self-contained ecosystem especially in locations where naturally occurring body of water is difficult to come by and land is limited and expensive. The objective of this research is; (i) To model a standard training database selection criterion on the fish behaviour in relation to their health, (ii) To develop an algorithm that can determine the health of the fish based on the behaviour of the fish, (iii) To implement and test the fish behaviour detection algorithm in a Biofloc fish pond to validate and evaluate the effectiveness of the system. The research methodology stages are; (i) Acquisition of daily recordings of the surfaces of fish ponds and their respective daily health, (ii)Formulating and developing a fish health monitoring algorithm based on their behaviour from the surface, (iii) Selection of training data based on the recordings and health,(iv) performance metric evaluation, (v) Assessment of experimental results of the Biofloc Fish Health Monitoring System On Pond Surface algorithm, and (vi) final report and project wrap-up. The expected outcome of this is an artificial intelligence system that can determine the well-being of the fishes in a Biofloc pond based on the behaviour of the fish that can be observed from the simulated environment of the pond. The proposed system can potentially reduce the already heavy workload of Biofloc fish farmers. |
|---|