激光与光电子学进展, 2019, 56 (6): 061101, 网络出版: 2019-07-30
基于深度学习和人眼视觉系统的遥感图像质量评价 下载: 1634次
Quality Assessment of Remote Sensing Images Based on Deep Learning and Human Visual System
成像系统 图像质量评价 深度学习 卷积神经网络 人眼视觉系统 imaging systems image quality assessment deep learning convolutional neural network human visual system
摘要
提出了一种基于深度学习和人眼视觉特性的遥感图像质量评价方法。利用卷积神经网络和反向传播神经网络分类器,同时对遥感图像进行特征学习及模糊和噪声强度的等级分类。利用掩盖效应和感知加权因子修正评价模型,得到了更符合人眼视觉的遥感图像质量评价结果。研究结果表明,所提方法有效解决了同时存在模糊和噪声的遥感图像质量评价的困难,能有效准确地评价遥感图像的质量,且与主观评价结果有较好的一致性,更符合人眼视觉感受。
Abstract
A quality assessment method of remote sensing images is proposed based on deep learning and the human visual characteristics. The convolutional neural network and the back propagation neural network classifiers are used for the simultaneous feature learning and grade classification of blur and noise intensity for the remote sensing images. The masking effect and the corrected assessment model of the perceptual weighting factors are used to obtain the quality assessment results of remote sensing images, which are more in line with the human visual characteristics. The research results show that the proposed method can effectively solve the difficulty in the quality assessment of remote sensing images with both blur and noise. Moreover, the quality of remote sensing images can be effectively and accurately evaluated, and the results are well in good agreement with both the subjective evaluation results and the human visual experiences.
刘迪, 李迎春. 基于深度学习和人眼视觉系统的遥感图像质量评价[J]. 激光与光电子学进展, 2019, 56(6): 061101. Di Liu, Yingchun Li. Quality Assessment of Remote Sensing Images Based on Deep Learning and Human Visual System[J]. Laser & Optoelectronics Progress, 2019, 56(6): 061101.