目的 为解决自然环境试验场中因地面不平导致智能巡检机器人位姿偏移,进而引发采集图像虚焦、歪斜等质量问题,并提升数据采集效率。方法 提出了一种集成机器人系统与智能算法的综合优化方法。首先,采用基于北斗卫星与语义地图的初次定位和基于数字标牌识别的二次定位相结合的双重定位机制,修正机器人行走轨迹偏差;其次,通过一种创新的边框扩张似物性检测算法对试验架整体图像进行精确划分,定位每个样品;最后,结合双目视觉原理动态计算最优拍摄参数（距离、角度、焦距）,确保目标垂直居中与清晰成像。结果 实验结果表明,经该方法优化后,机器人采集的图像质量显著提升,自然图像质量评估（NIQE）分数从4.91优化至4.03,图像合格率从75%提升至95%,与人工采集相当。同时,数据采集频率从人工的每3个月1次提升至每日1次,效率提升约90倍。结论 研究构建了一套自动化、高精度的图像采集方案,有效克服了人工采集和传统机器人巡检的局限性,显著提升了自然环境试验数据采集的时效性、规范性与科研价值,为装备环境效应研究提供了可靠的技术支持。
Abstract
The work aims to address the issues of image defocusing, tilting, and other quality problems caused by pose shifts of intelligent inspection robots due to uneven ground in natural environmental test sites and enhance data acquisition efficiency. An integrated optimization method combining a robotic system with intelligent algorithms was proposed. Firstly, a dual localization mechanism integrating initial positioning based on the BeiDou satellite and semantic map with secondary positioning based on digital tag recognition was adopted to correct the robot's walking trajectory deviation. Subsequently, an innovative border-expanding objectness detection algorithm was used to accurately segment the overall image of the test rack and locate each sample. Finally, combined with the principle of binocular vision, the optimal shooting parameters (distance, angle, focal length) were dynamically calculated to ensure that the target was vertically centered and clearly imaged. Experimental results indicated that after optimization by this method, the image quality captured by the robot significantly improved. The Natural Image Quality Evaluator (NIQE) score was optimized from 4.91 to 4.03, and the image qualification rate increased from 75% to 95%, which was comparable to manual acquisition. Meanwhile, the data acquisition frequency increased from once every three months manually to once per day, representing an approximately 90-fold efficiency improvement. This work establishes an automated, high-precision image acquisition scheme that effectively overcomes the limitations of manual acquisition and traditional robot inspection, which significantly enhances the timeliness, standardization, and scientific value of data acquisition in natural environmental tests, providing reliable technical support for research on equipment environmental effects.
关键词
智能巡检机器人 /
环境试验 /
数据采集 /
图像质量优化 /
机器视觉
Key words
intelligent inspection robot /
environmental test /
data acquisition /
image quality optimization /
machine vision
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] XIANG L, TAO J Q, XIA X S, et al.Impact of Marine Atmospheric Corrosion on the Microstructure and Tensile Properties of 7075 High-Strength Aluminum Alloy[J]. Materials, 2023, 16(6): 2396.
[2] 张伦武, 周堃, 赵方超, 等. 装备环境适应性研究进展及展望[J]. 装备环境工程, 2024, 21(5): 1-12.
ZHANG L W, ZHOU K, ZHAO F C, et al.Research Progress and Prospect of Materiel Environmental Worthiness[J]. Equipment Environmental Engineering, 2024, 21(5): 1-12.
[3] SHEN Y F, MA R Y, WEI J, et al.Influence of Sea Sand and Shewanella Algae on the Corrosion Behavior of 316 Stainless Steel in Marine Environment[J]. Corrosion Science, 2024, 235: 112200.
[4] CLARK B, ALLEN M S, PACINI B.Study of the Effect of Bolted Joint Damping-Stiffness Nonlinearity on Part Failures in Dynamic Environment Testing[J]. Experimental Techniques, 2026, 50(1): 163-174.
[5] WANG B Q, LIU L G, CHENG X Q, et al.Advanced Multi-Image Segmentation-Based Machine Learning Modeling Strategy for Corrosion Prediction and Rust Layer Performance Evaluation of Weathering Steel[J]. Corrosion Science, 2024, 237: 112334.
[6] HORIE T, KITAHARA G, TANI H, et al.Changes in Admittance and Internal Structure of Coating Films by Environmental Testing[J]. Materials Transactions, 2024, 65(10): 1287-1292.
[7] 覃粒, 吴德权, 胡涛, 等. 海南湿热海洋大气环境Q235钢腐蚀行为研究及严酷度评估[J]. 装备环境工程, 2023, 20(7): 90-97.
QIN L, WU D Q, HU T, et al.Corrosion Behaviors of Q235 Steel and Severity Evaluation for Humid and Hot Marine Atmosphere Environmental in Hainan[J]. Equipment Environmental Engineering, 2023, 20(7): 90-97.
[8] 王荣祥, 龚雨荷, 于宏飞, 等. 飞机典型结构连接件自然环境试验方法研究[J]. 环境技术, 2024, 42(8): 90-94.
WANG R X, GONG Y H, YU H F, et al.Study on the Natural Environment Test Method of Typical Aircraft Structural Connections[J]. Environmental Technology, 2024, 42(8): 90-94.
[9] 陈星昊, 舒畅, 黄伦, 等. 环境效应检测与评价中的图像识别技术综述[J]. 兵器装备工程学报, 2021, 42(12): 44-51.
CHEN X H, SHU C, HUANG L, et al.Image Recognition Technology in Detection and Evaluation of Environmental Effect[J]. Journal of Ordnance Equipment Engineering, 2021, 42(12): 44-51.
[10] 黄伦, 朱玉琴, 舒畅, 等. 智能巡检机器人在自然环境试验中的应用探讨[J]. 装备环境工程, 2024, 21(1): 121-126.
HUANG L, ZHU Y Q, SHU C, et al.Discussion on Application of Intelligent Inspection Robots in Natural Environment Experiments[J]. Equipment Environmental Engineering, 2024, 21(1): 121-126.
[11] PREETHICHANDRA D M G, PIYATHILAKA L, IZHAR U. Review on Robotic Systems for Environmental Monitoring[J]. IEEE Open Journal of Instrumentation and Measurement, 2025, 4: 9500317.
[12] LI D J, SHI H T, CAI B, et al.A Review of Technical Advances and Applications of Intelligent Inspection Robots in Structural Health Monitoring[J]. SmartBot, 2025, 1(3): e70000.
[13] LI D Y, ZHONG Z F, CHEN T X, et al.Research on Optimization of Complex Path of Inspection Robot[J]. Journal of Physics: Conference Series, 2021, 2029(1): 012124.
[14] MENG Q L, YANG J B, ZHANG Y, et al.A Robot System for Rapid and Intelligent Bridge Damage Inspection Based on Deep-Learning Algorithms[J]. Journal of Performance of Constructed Facilities, 2023, 37(6): 04023052.
[15] DAMJANOVIĆ D, BIOČIĆ P, PRAKLJAČIĆ S, et al. A Comprehensive Survey on SLAM and Machine Learning Approaches for Indoor Autonomous Navigation of Mobile Robots[J]. Machine Vision and Applications, 2025, 36(3): 55.
[16] GUAN J W.Robot SLAM Method Based on Multi-Sensor Fusion[J]. Journal of Engineering System, 2024, 2(2): 84-88.
[17] DU H, WANG X D, WANG P F, et al.Image Super-Resolution Reconstruction for Corrosion Damage of Coated Steel Components[J]. Construction and Building Materials, 2025, 494: 143412.
[18] CUI S H, HU Z Y, QIU F C, et al.Fault Detection for Split Pins of Power Transmission Fittings in UAV Inspections via Automatic Image Cropping-Based Super-Resolution Reconstruction and Enhanced YOLOv8[J]. Journal of Electronic Research and Application, 2025, 9(3): 222-234.
[19] ZHOU J P, XUE L T, LI Y, et al.A Novel Fuzzy Controller for Visible-Light Camera Using RBF-ANN: Enhanced Positioning and Autofocusing[J]. Sensors, 2022, 22(22): 8657.
[20] KANG H, PARK G.Development of YOLO-Based Object Detection Model with Bounding Box Separation Techniques[J]. International Journal of Automotive Technology, 2026, 27(2): 843-856.
[21] GUO Y C, DU L, LYU G X.SAR Target Detection Based on Domain Adaptive Faster R-CNN with Small Training Data Size[J]. Remote Sensing, 2021, 13(21): 4202.
基金
国家自然科学基金（U22A20101）; 重庆市技术创新与应用发展专项重点项目（CSTB2022TIAD-CUX0015）
PDF(1249 KB)
