[1]鲍克鹏,许 涛,贾雪磊,等.基于改进双边分割网络的视觉测振方法[J].机械与电子,2025,(03):29-34.
　BAO Kepeng,XU Tao,JIA Xuelei,et al.Visual Vibration Measurement Method Based on an Improved Bilateral Segmentation Network[J].Machinery & Electronics,2025,(03):29-34.
点击复制

基于改进双边分割网络的视觉测振方法()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
期数:
2025年03期
页码:
29-34
栏目:
自动控制与检测
出版日期:
2025-03-25

文章信息/Info

Title:
Visual Vibration Measurement Method Based on an Improved Bilateral Segmentation Network
文章编号:
1001-2257 ( 2025 ) 03-0029-06
作者:
鲍克鹏; 许 涛; 贾雪磊; 王 栋
西安工程大学机电工程学院,陕西 西安 710600
Author(s):
BAO Kepeng ;  XU Tao ;  JIA Xuelei ;  WANG Dong
( School of Mechanical and Electrical Engineering , Xi ’ an Polytechnic University , Xi ’ an 710600 , China )
关键词:
计算机视觉; 语义分割; 注意力机制; 多尺度特征融合
Keywords:
computer vision ;  semantic segmentation ;  attention mechanism ;  multi-scale feature fusion
分类号:
TP391.4
文献标志码:
A
摘要:
为解决双边分割网络在标签分割时存在的分割边界拟合精度低以及边缘不准确而导致的形心计算偏差问题,提出一种改进的 BiSeNet 模型。首先,改进损失监测机制,提升模型对目标对象的关注度;其次,设计了一种多尺度边缘融合模块,增强网络的特征表示和训练稳定性。消融实验和对比实验结果表明,改进的双边分割网络分割模型,其 mIoU 和 mF1 评价指标较基准网络分别提高了 2.82 百分点和 2.51百分点。为验证其在振动测量中的有效性,以振动实验台作为实验对象。实验结果显示,与传统加速度传感器相比,所提方法测得的振动数据与真实位移数据之间的均方误差仅为 0.06 。
Abstract:
In order to solve the problem of centroid calculation deviation caused by low fitting accuracy and inaccurate edge of two sided segmentation network in label segmentation.An improved BiSeNet model is proposed.Firstly , the loss monitoring mechanism is improved to enhance the model’s attention to the target object ; secondly , a multi scale edge fusion module is designed to enhance the feature representation and training stability of the network.The results of ablation experiment and comparison experiment show that the mIoU and mF1 evaluation indexes of the improved bilateral segmentation network are 2.82 per- centage points and 2.51 percentage points higher than the benchmark network , respectively.In order to verify its effectiveness in vibration measurement , the vibration test bench is taken as the experimental object in this paper.The experimental results show that the mean square error between the measured vibration data and the real displacement data is only 0.06 compared with the traditional acceleration sensor.

参考文献/References:

[ 1 ] BAQERSAD J , POOZESH P , NIEZRECKI C , et al.Photogrammetry and optical methods in structural dynamics A review [ J ] .Mechanical systems and signal processing , 2017 , 86 : 17-34.

[ 2 ] FENG D M , FENG M Q.Computer vision for SHM of civil infrastructure : from dynamic response measurement to damage detection A review [ J ] .Engineering structures , 2018 , 156 : 105-117.
[ 3 ] WENG J H , CHEN L , SUN L M , et al.Fully automated and non-contact force identification of bridge cables using microwave remote sensing [ J ] .Measurement , 2023 , 209 : 112508.
[ 4 ] 董衡 . 差分干涉雷达形变测量关键技术研究[ D ] . 长沙:国防科学技术大学,2019.
[ 5 ] KALPOE D , KHOSHELHAM K , GORTE B.Vibration measurement of a model wind turbine using high speed photogrammetry [ C ] ∥Videometrics , range imaging , and applications XI , 2011 : 80850J.
[ 6 ] 邵新星,陈振宁,戴云彤,等 . 数字图像相关方法若干关键问题研究进展 [ J ] . 实验力学, 2017 , 32 ( 3 ):305-325.
[ 7 ] 钱荣威,马增强,许丹丹,等 . 基于改进DOG边缘检测算子的铁鞋相对位移检测[ J ] . 组合机床与自动化加工技术,2021 ( 3 ): 94-97.
[ 8 ] 苏磊,黄华,秦伦明,等 . 基于 YOLOv4 及金字塔 LK光流法的变压器振动检测方法[ J ] . 广东电力, 2022 , 35( 9 ): 9-18.
[ 9 ] 王浩林,刘源,王芸 . 基于相位光流法的结构振动微小位移测量 [ J ] . 计算机系统应用, 2022 , 31 ( 2 ):358-365.
[ 10 ] 陈俊松,龚俊,张格悠,等 . 基于改进 L K 光流法的电缆断裂伸长率测量技术[ J ] . 仪器仪表学报, 2021 ,42 ( 4 ): 256-264.
[ 11 ] 冯海龙,刘伯奇,胡海天,等 . 基于视觉测振技术的高铁无站台柱雨棚振动测量方法[ J ] . 地震工程与工程振动,2022 , 42 ( 3 ): 209-215.
[ 12 ] CUADRADO M , PERNAS S?NCHEZ J , ARTERO GUERRERO J A , et al.Model updating of uncertain parameters of carbon / epoxy composite plates using digital image correlation for full-field vibration measurement [ J ] .Measurement , 2020 , 159 : 107783.
[ 13 ] 晏班夫,欧阳康,梁才 . 桥梁工程中非接触位移测量技术研究综述 [ J ] . 交通运输工程学报,2024 , 24 ( 1 ):43-67.
[ 14 ] 赵明,唐淋,陈石,等 . 基于深度学习到时拾取自动构建长宁地震前震目录[ J ] . 地球物理学报,2021 , 64( 1 ): 54-66.
[ 15 ] 陈立潮,解丹,曹建芳,等 . 改进光流法和 GMM 融合的车辆实时检测算法研究[ J ] . 智能系统学报, 2021 ,16 ( 2 ): 271-278.
[ 16 ] SAXENA D , CAO J N.Generative adversarial networks ( GANs ) challenges , solutions , and future directions [ J ] .ACM Computing surveys ( CSUR ), 2021 , 54 ( 3 ):1-42.
[ 17 ] ZHANG Y , LIU P , ZHAO X F.Structural displacement monitoring based on mask regions with convolutional neural network [ J ] .Construction and building materials , 2021 , 267 : 120923.
[ 18 ] CHAI S L , WANG S , LIU C , et al.A visual measurement algorithm for vibration displacement of rotating body using semantic segmentation network [ J ] .Expert systems with applications , 2024 , 237 : 121306.

相似文献/References:

[1]陈 红,胡恩才,白静静.交叉路口信号灯智能管理系统设计[J].机械与电子,2019,(11):63.
　,Design of Intelligent Management System for Intersection Signal Lights[J].Machinery & Electronics,2019,(03):63.
[2]郑卓纹,吴攀超,王婷婷,等.基于轻量化目标检测算法的指针仪表读数识别[J].机械与电子,2025,(05):10.
　ZHENG Zhuowen,WU Panchao,WANG Tingting,et al.Pointer Gauge Reading Recognition Based on Lightweight Target Detection Algorithm[J].Machinery & Electronics,2025,(03):10.
[3]李晓亮,李光亚,孟志琳. 基于Swin-UNet的破损碑刻文字识别方法[J].机械与电子,2026,44(01):28.
　LI Xiaoliang,LI Guangya,MENG Zhilin. A Character Recognition Method for Damaged Inscriptions Based on Swin-UNet[J].Machinery & Electronics,2026,44(03):28.

备注/Memo

备注/Memo:
收稿日期: 2024-09-05
基金项目:陕西省自然科学基础研究计划( 2024JC-YBMS-403 );陕西省技术创新引导计划( 2024QCY-KXJ-030 )
作者简介:鲍克鹏 ( 1996- ),男,甘肃张掖人,硕士研究生,研究方向为基于机器视觉的振动测量;许 涛 ( 1983- ),男,陕西渭南人,副教授,硕士研究生导师,研究方向为数字信号和图像处理,通信作者;贾雪磊 ( 2000- ),男,安徽宿州人,硕士研究生,研究方向为计算机视觉振动测量;王 栋 ( 1999- ),男,甘肃张掖人,硕士研究生,研究方向为计算机视觉振动测量。
更新日期/Last Update: 2025-04-07