Page 110 - 《应用声学》2023年第3期

第 42 卷 第 3 期 Vol. 42, No. 3
2023 年 5 月 Journal of Applied Acoustics May, 2023

⋄ 研究报告 ⋄

基于Rician分布散斑噪声的超声图像模拟算法 ∗

朋小秀 张 东 †

(武汉大学物理科学与技术学院 武汉 430072)

摘要：为了衡量各种去噪算法的性能，在干净无噪声的图像上添加接近真实且可控的散斑噪声是非常重要的。
提出一种基于 Rician 分布的不完全发育的斑点噪声的超声图像模拟算法。该算法考虑到了声波区域中包含孤
立的强散射体的情况，同时结合了超声成像的扫描过程。以合成图像和肾脏图像为体模进行了模拟实验，并对
最终生成的伪超声图像进行了噪声分布统计及拟合检验。实验结果表明该算法生成的伪超声图像在视觉上和
理论上都接近真实的超声图像。
关键词：超声图像模拟；合成图像；Rician 分布；分布拟合检验
中图法分类号: TB552 文献标识码: A 文章编号: 1000-310X(2023)03-0548-10
DOI: 10.11684/j.issn.1000-310X.2023.03.013

Ultrasound image simulation algorithm based on speckle noise based on
Rician distribution

PENG Xiaoxiu ZHANG Dong

(School of Physics and Technology, Wuhan University, Wuhan 430072, China)

Abstract: In order to measure the performance of various denoising algorithms, it is very important to add
near-real and controllable speckle noise to clean and noise-free images. An ultrasonic image simulation algo-
rithm of incompletely developed speckle noise based on Rician distribution is proposed. The algorithm takes
into account the fact that there are isolated strong scatterers in the acoustic region, and combines with the
ultrasound imaging scanning process. The synthetic image and the kidney image are used as phantoms for
simulation experiments, and the noise distribution statistics and fitting tests are carried out on the final gener-
ated pseudo-ultrasound images. The experimental results show that the pseudo-ultrasound images generated
by the algorithm are close to the real ultrasound images both visually and theoretically.
Keywords: Ultrasound image simulation; Synthetic image; Rician distribution; Distribution fitting test

2022-01-27 收稿; 2022-03-28 定稿
国家重点研发计划项目 (2011CB707900)
∗
作者简介: 朋小秀 (1996– ), 女, 安徽人, 硕士, 研究方向: 超声图像模拟。
† 通信作者 E-mail: dz_whu@163.com