# SAR Image Despeckling Using a Convolutional Neural Network

## TL;DR

This paper introduces ID-CNN, a deep learning model that effectively removes speckle noise from SAR images, improving image quality for better interpretation and processing.

## Contribution

The paper presents a novel end-to-end CNN architecture with a division residual layer for SAR despeckling, outperforming existing methods.

## Key findings

- Significant noise reduction on synthetic SAR images.
- Improved image clarity on real SAR data.
- Outperforms state-of-the-art despeckling techniques.

## Abstract

Synthetic Aperture Radar (SAR) images are often contaminated by a multiplicative noise known as speckle. Speckle makes the processing and interpretation of SAR images difficult. We propose a deep learning-based approach called, Image Despeckling Convolutional Neural Network (ID-CNN), for automatically removing speckle from the input noisy images. In particular, ID-CNN uses a set of convolutional layers along with batch normalization and rectified linear unit (ReLU) activation function and a component-wise division residual layer to estimate speckle and it is trained in an end-to-end fashion using a combination of Euclidean loss and Total Variation (TV) loss. Extensive experiments on synthetic and real SAR images show that the proposed method achieves significant improvements over the state-of-the-art speckle reduction methods.