An Underwater Image-Based Object Classifier Using Convolutional Neural Network
Hena Prince, T. Binesh · 2024
Underwater images are characterized by low-quality issues due to poor illumination, attenuation, and scattering of light in the under-sea. So underwater object classification based on images is a challenging task in itself. The proposed methodology presents a convolutional neural network for under-sea object classifications and the results are compared with the Feedforward Fully Connected Network(FFCN). The proposed classifier exhibits an accuracy of 81% and the evaluation metrics also reveal the improved efficiency of the proposed method.