【24h】

SRNet: A Cascade Network to Speckle Reduction of SAR Image

机译:SRNET:级联网络到SAR图像的散斑减少

获取原文
获取外文期刊封面目录资料

摘要

Speckle noise limits the usage of synthetic aperture radar (SAR) for object recognition and segmentation tasks.Most traditional methods sacrifice useful image information to achieve speckle reduction. The classic method based onlocal sliding window filtering has obvious side effect of erasing object edges and blurring texture information comparingwith ground truth image. Another widely used method is convolutional neural network based on mean squared error, thevisual effect of denoised image is not satisfactory even though MSE loss can have higher peak signal-to-noise ratio(PSNR) performance. In this paper, we present a cascade network to address this problem, namely SRNet, whichemploys an asymmetric architecture for the task of speckle noise reduction. The cascade architecture can supervise thenetwork to revise on both pixel-wise level and feature-wise level by calculating correlation coefficient loss on the featuremaps. In the meanwhile, we utilize the auxiliary loss on the intermediate results to accelerate the convergence of thenetwork. The proposed network preserves the edge texture details much better than other compared methods.
机译:散斑噪声限制了合成孔径雷达(SAR)的使用进行对象识别和分割任务。大多数传统方法牺牲有用的图像信息以实现散斑减少。基于的经典方法局部滑动窗滤波具有擦除物体边缘和模糊纹理信息比较的明显副作用与地面真相形象。另一种广泛使用的方法是基于均方误差的卷积神经网络,即使MSE损耗可能具有较高的峰值信噪比,即使MSE损失也可能具有较高的峰值信噪比,去噪图像的视觉效果也不令人满意(PSNR)性能。在本文中,我们提出了一个级联网络来解决这个问题,即Srnet,哪个采用非对称架构,用于减少斑块降噪任务。级联架构可以监督通过计算特征的相关系数损耗来修改两个像素 - WISE级别和特征方面的网络地图。同时,我们利用中间结果的辅助损失来加速趋同网络。所提出的网络保留了优于其他比较方法的边缘纹理细节。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号