一种融合NSCT-HMT 与FLICM 的矿山遥感图像去噪算法#br#

金属矿山 ›› 2026, Vol. 55 ›› Issue (5): 296-302.

一种融合NSCT-HMT 与FLICM 的矿山遥感图像去噪算法#br#

李　琳^1,2　杨泽辉¹　魏　巍³　陈　伟⁴　唐　强⁴

1. 太原理工大学财经学院,山西太原 030024;2. 山西省财政税务专科学校信息科技学院,山西太原 030024;
3. 山西大学计算机与信息技术学院,山西太原 030006;4. 苏州大学轨道交通学院,江苏苏州 215131

出版日期:2026-05-15 发布日期:2026-06-04
通讯作者: 唐　强(1985—),男,教授,博士,博士研究生导师。
作者简介:李　琳(1977—),女,副教授,硕士。
基金资助:
“十四五”国家重点研发计划项目(编号:2023YFC3707801);山西省高等学校科技创新项目(编号:2024L541);国家自然科学基金项目
(编号:62276160,52478352)。

A Noise Reduction Algorithm for Mining Remote Sensing Images Integrating NSCT-HMT and FLICM

LI Lin^1,2　YANG Zehui¹　WEI Wei³　CHEN Wei⁴　TANG Qiang⁴

1. College of Finance and Economics,Taiyuan University of Technology,Taiyuan 030024,China;
2. Information Technology Institute,Shanxi Finance & Taxation College,Taiyuan 030024,China;
3. School of Computer and Information Technology,Shanxi University,Taiyuan 030006,China;
4. School of Rail Transportation,Soochow University,Suzhou 215131,China

Online:2026-05-15 Published:2026-06-04

摘要/Abstract

摘要： 矿山遥感图像在资源调查与灾害预警中具有重要作用,但受复杂成像环境和散射效应影响,常伴随严
重噪声与纹理畸变,影响了识别精度。为此,提出了一种融合非下采样轮廓波变换(Nonsubsampled Contourlet Transform,
NSCT)与隐马尔可夫树(Hidden Markov Tree,HMT)及模糊局部C-均值(Fuzzy Local Information C-Means,FLICM)
的矿山遥感图像去噪算法。首先利用非下采样轮廓波变换实现多尺度分解,再以隐马尔可夫树实现父子系数的跨尺
度依赖关系建模;随后结合模糊局部C 均值聚类与图卷积网络强化空间邻域和跨邻域特征聚合,从而实现对复杂噪
声的有效抑制与细节特征的保持。试验结果表明:该算法在UCM、AID 公开数据集上峰值信噪比(Peak Signal-to-Noise
Ratio,PSNR)达到34. 59、33. 74 dB,结构相似度(Structural Similarity Index,SSIM)为0. 917 和0. 904,梯度相似偏差
(Gradient Magnitude Similarity Deviation,GMSD)为0. 072 和0. 082,单帧操作延迟为35. 18、36. 29 ms,性能优于生成对
抗网络(Generative Adversarial Network,GAN)、Transformer 以及混合自监督等算法。研究反映出,该算法在复杂矿山环
境下具备较高的鲁棒性与应用价值,为矿山遥感图像去噪与灾害隐患识别提供了便利。

关键词: 矿山遥感图像　, 图像去噪　, NSCT　, HMT　, FLICM

Abstract: Mine remote sensing images play a crucial role in resource surveying and disaster warning. However,due to
complex imaging environments and scattering effects,they often suffer from severe noise and texture distortion,which compromises
recognition accuracy. To address this issue,a denoising algorithm is proposed for mine remote sensing images by integrating
the Nonsubsampled Contourlet Transform (NSCT) with the Hidden Markov Tree (HMT) and Fuzzy Local Information CMeans
(FLICM). Firstly,NSCT is employed to perform multi-scale decomposition. Then,the HMT is used to model cross-scale
dependencies between parent and child coefficients. Subsequently,FLICM clustering is combined with a Graph Convolutional
Network (GCN) to enhance the aggregation of both spatial neighborhood and cross-neighborhood features,thereby effectively
suppressing complex noise while preserving fine details. Experimental results on the UCM and AID public datasets show that
the proposed algorithm achieves a Peak Signal-to-Noise Ratio (PSNR) of 34. 59 dB and 33. 74 dB,a Structural Similarity Index
(SSIM) of 0. 917 and 0. 904,and a Gradient Magnitude Similarity Deviation (GMSD) of 0. 072 and 0. 082,respectively.
The per-frame processing latency is 35. 18 ms and 36. 29 ms. The performance outperforms that of Generative Adversarial Network
(GAN),Transformers,and hybrid self-supervised methods. The study demonstrates that the proposed algorithm possesses
high robustness and practical value in complex mine environments,facilitating mine remote sensing image denoising and hazard identification.

Key words: mine remote sensing image,image denoising,NSCT,HMT,FLICM

中图分类号:

TD67
TP391

李　琳, 杨泽辉, 魏　巍, 陈　伟, 唐　强. 一种融合NSCT-HMT 与FLICM 的矿山遥感图像去噪算法#br#[J]. 金属矿山, 2026, 55(5): 296-302.

LI Lin, YANG Zehui, WEI Wei, CHEN Wei, TANG Qiang. A Noise Reduction Algorithm for Mining Remote Sensing Images Integrating NSCT-HMT and FLICM[J]. Metal Mine, 2026, 55(5): 296-302.

[1]	贾亚娟　戴二壮　王咏宁. 适用于矿山遥感图像增强的改进对抗生成网络算法[J]. 金属矿山, 2025, 54(9): 272-278.
[2]	沈丹萍, 赵　爽. 基于轻量化YOLOX-S 与多阈值分割的矿山遥感图像去噪算法[J]. 金属矿山, 2024, 53(9): 175-.
[3]	赵　杰, 汪洪法, 周　明. 融合眼视觉和改进多尺度Retinex 的矿岩裂隙图像增强[J]. 金属矿山, 2024, 53(12): 252-.

扫码分享