Sar Ship Detection Using Feature Enhancement ...

Sar Ship Detection Using Feature Enhancement Pyramid (FEP) and Shallow Feature Reconstruction (SFR) Networks In Mat Lab

Author Name : Mr. B. Malakonda Reddy, A. Surendra, K. Bharath Kumar, K. Gowtham Babu, K. Siva Reddy

ABSTRACT The feature enhancement pyramid and shallow feature reconstruction network (FERSFRN) is a deep learningbased approach proposed for the SAR Ship Detection project. The FER-SFRN method aims to improve the accuracy and efficiency of ship detection in SAR images by enhancing the features in the images and reconstructing shallow features. The FER-SFRN method is based on a pyramid structure that enhances the features in SAR images at multiple scales. The pyramid structure consists of a feature enhancement module (FEM) and a shallow feature reconstruction module (SFRM). The FEM module enhances the features in SAR images by using a convolutional neural network (CNN) to extract features and a feature pyramid module to enhance the features at multiple scales. The SFRM module reconstructs the shallow features by using a shallow feature reconstruction network that maps the enhanced features back to the original image space.