CAE-CNN Hybrid Model for Efficient Classification of Freshwater Fish Diseases

Detecting fish infections and monitoring their health via manual visual examination is a difficult undertaking. To avoid disease transmission, diseased fish must be identified as soon as possible. As a result, tackling this problem requires a rapid, dependable, and highly automated solution. This work presents a novel hybrid deep-learning architecture for the classification of freshwater fish diseases. The proposed model leverages the strengths of both Convolutional Autoencoders (CAEs) and Convolutional Neural Networks (CNNs) to achieve improved accuracy and efficiency in disease identification. First, a CAE is used to reduce dimensionality and capture complex illness patterns by extracting significant characteristics from fish images. After that, a CNN is fed the encoded features, and it completes the last classification task. This hybrid method makes use of CNN’s expertise in picture classification and the CAE’s capacity to learn discriminative representations. Experimental findings show that the suggested design outperforms conventional CNN-only models, indicating its potential for precise and prompt identification of diseases affecting freshwater fish. This technique has the potential to greatly improve aquaculture health management by facilitating timely treatment and reducing possible financial losses.