Underwater Debris Detection and Classification Using Deep Learning Models

Abstract

Debris present inside water poses significant impacts on under water ecosystems along with living organisms thriving in them. In order to address the issue of water pollution and environmental hazards, detection and removal of this growing underwater debris is utmost need of the present times. Therefore, this study explores the application of deep neural network model specifically pretrained vgg16 for detection and classification of underwater debris. Additionally, the current study presents a comprehensive locally collected under water debris images dataset for the detection and classification of debris in local underwater environment is also proposed. The proposed custom vgg16 model performs well in detection and classification of underwater debris with an accuracy of 84%. Moreover, this model is effectively proficient in detecting plastic debris present inside water environment. Furthermore, the model's strength was authenticated through testing on unseen underwater debris images, showcasing its image detection potential for real underwater ecosystem deployment. This study adds to the progression of automatic underwater detection systems, proposing a viable tool for environmental mitigation.