The purpose of investigation and adjudication on marine accidents is to contributeto securing maritime safety by identifying the causes of maritime accidents andpreventing the recurrence of similar maritime accidents. Recently, movements suchas scient...
The purpose of investigation and adjudication on marine accidents is to contributeto securing maritime safety by identifying the causes of maritime accidents andpreventing the recurrence of similar maritime accidents. Recently, movements suchas scientific analysis methods and accident reproduction and verification based onaccident site data to analyze the cause of marine accidents and prevent recurrenceare considered very positive for accident investigation and establishment ofaccountability system. Therefore, this study proposes a lightweight neural network,the ULSR model, to solve the complexity problem of the underwater imagesuper-resolution neural network model to support the ultimate purpose of thisaccident investigation. The core of the proposed idea is to integrate the same modelsize for all super-resolution magnifications, configure it to have similar number ofparameters and inference speed, and achieve high PSNR and SSIM results at allmagnifications compared to the existing model.