Fish Disease Detection Using Image Based Machine Learning Technique in Aquaculture

The increasing demand for fish production in aquaculture has highlighted the need for efficient disease detection systems to maintain the health of aquatic species. This study proposes a machine learning-based approach for fish disease detection using image processing techniques. A dataset comprising images of both fresh and infected fish is sourced from online repositories. The images undergo pre-processing steps such as resizing, contrast enhancement, and color space conversion to standardize and improve feature extraction for optimal classification performance. The pre-processed images are then fed into a Support Vector Machine (SVM) classifier, a robust machine learning algorithm, to distinguish between fresh and infected fish. The system classifies the input images into two categories: "Fresh" or "Infected Fish." If the system detects infected fish, it highlights the affected regions in red to provide visual insights into the infection's location. The proposed approach is evaluated for accuracy to ensure reliable real-world performance. This automated system offers a more efficient and accurate alternative to manual inspections, supporting timely disease detection and management in aquaculture. By enabling early identification of infections, the system contributes to sustainable fish farming practices and reduces the risk of disease spread, ultimately supporting the growth of healthy fish populations.

Keywords: Fish Dataset, Pre-Processing, Support Vector Machine, Machine learning, Classification, Accuracy.