基于NSST和深度学习的医学图像融合算法
Medical Image Fusion Algorithm Based on NSST and Deep Learning
郝昱权 1明子琚1
作者信息
- 1. 驻马店幼儿师范高等专科学校,河南驻马店 463000
- 折叠
摘要
随着计算机性能的提高和数字图像处理技术的发展,医学图像处理技术得到了更广泛的应用,如医学图像的自动分析、医学图像的模拟仿真等.因此提出一种基于NSST与VGGNet19深度学习网络相结合的医学图像融合方法.首先,对医学源图像进行NSST分解,分别得到MRI与CT的低频信息与高频信息.其次,低频信息选择引导滤波加权均等的融合规则,高频信息用VGGNet19网络对图像进行特征提取,通过L1正则化、上采样以及加权均等的规则得到最终的高频信息.实验结果表明该方法有更好的融合效果,在主观评价和客观指标中均有不错的效果体现.
Abstract
With the improvement of computer performance and the development of digital image processing technology,the medical image processing technology has been more widely used,such as the automatic analysis of medical images,the simulation of medical images,etc. Therefore,a medical image fusion method based on combining NSST and VGGNet19 deep learning network is proposed. First,the NSST decomposition of the medical source images was performed to obtain the low-frequency information and the high-frequency information of MRI and CT,respectively. Secondly,the low frequency information selects the fusion rule of guided filter and weighted equalization. The high frequency information extracts the image features with VGGNet19 network,and the final high frequency information is obtained through the rules of L1 regularization,upsampling and weighted equalization. The experimental results show that this method has a better fusion effect,which is good reflected in both subjective evaluation and objective indicators.
关键词
非下采样剪切波/深度学习网络/图像融合Key words
non-subsampled shearlet transform/deep learning network/image fusion引用本文复制引用
基金项目
2024年度驻马店幼儿师范高等专科学校校级课题(24L010)
出版年
2024
NETL
NSTL
万方数据