基于人工智能技术的激光快速测量系统误差检测  

Error detection of laser rapid measurement system based on artificial intelligence technology

作　　者：谷栎娜 薛少童 张萌 沈娴[1] GU Li’na;XUE Shaotong;ZHANG Meng;SHEN Xian(Artificial intelligence and big date college,Hebei University of Engineering Science,Shijiazhuang 050091,China;Department of electronic information,Huaxin College of Hebei GEO University,Shijiazhuang 050700,China)

机构地区：[1]河北工程技术学院人工智能与大数据学院,石家庄050091 [2]河北地质大学华信学院电子信息系,石家庄050700

出　　处：《激光杂志》2025年第4期97-102,共6页Laser Journal

基　　金：河北省高等学校科学技术研究项目(No.ZD2020410)。

摘　　要：在实际应用过程中,由于测量设备本身的精度限制、测量环境的干扰和人为操作等因素影响,激光快速测量系统不可避免地会存在一定的误差。误差的存在不仅会影响测量结果的准确性,还可能对后续的数据分析和决策制定产生误导。因此,设计基于人工智能技术的激光快速测量系统误差检测方法。首先对激光快速测量系统成像实施畸变校正处理。然后利用改进的AHE算法和双线性插值法对图像进行增强处理,进一步改善激光快速测量系统成像的质量。最后选用经过优化的VGG网络作为本技术的卷积神经网络架构。利用该单卷积神经网络拟合图像与误差之间的映射关系,检测出该系统的误差。实验测试结果表明,设计方法检测到的实验系统检测误差与实际误差一致。In practical applications,due to the accuracy limitations of the measuring equipment itself,interference from the measuring environment,and human operations,laser rapid measurement systems inevitably have certain errors.The existence of errors not only affects the accuracy of measurement results,but may also mislead subsequent data analysis and decision-making.Therefore,a laser rapid measurement system error detection method based on artificial intelligence technology is designed.Firstly,distortion correction is applied to the imaging of the laser rapid measurement system.Then,the improved AHE algorithm and bilinear interpolation method are used to enhance the image and further improve the imaging quality of the laser rapid measurement system.Finally,the optimized VGG network was selected as the convolutional neural network architecture for this technology.Use the single convolutional neural network to fit the mapping relationship between the image and the error,and detect the error of the system.The experimental test results show that the detection error of the experimental system detected by the design method is consistent with the actual error.

关 键 词：人工智能技术 激光快速测量系统 切向畸变模型 VGG网络 误差检测 

分 类 号：TN929[电子电信—通信与信息系统]