基于数据挖掘的数字绘本图像传感器故障自动诊断方法
Automatic Fault Diagnosis Method for Digital Picture Book Image Sensors Based Data Mining
南姣鹏1
作者信息
摘要
图像传感器作为多个行业的重要检测工具,通常被运用于图像信号的采集和信息获取.但在恶劣的工作环境中,传感器通常会出现精度下降等问题,因此此次研究为了提升图像传感器的故障检测精度,将数据挖掘技术引入传统的故障检测识别算法和故障诊断算法中进行算法的改进.研究结果表明,改进后的检测识别算法精度比传统算法高11.1％,运算速度更快;改进后的故障诊断算法对故障诊断准确度明显提升,召回率也相对较高;算法模型在信噪比相同时还具备更高的故障识别准确率,同时在故障识别时的训练时间更短.因此,改进后的算法性能明显提高,在数字绘本图像传感器故障诊断中具有重要意义.
Abstract
As an important detection tool in multiple industries,image sensors are usually used for image signal acquisition and information acquisition.However,in harsh working environments,sensors often experience issues such as decreased accuracy.There-fore,in order to improve the fault detection accuracy of image sensors,this study introduces data mining technology into traditional fault detection and recognition algorithms and fault diagnosis algorithms for algorithm improvement.The research results show that the improved detection and recognition algorithm has an accuracy 11.1％higher than traditional algorithms and a faster computational speed;The improved fault diagnosis algorithm significantly improves the accuracy of fault diagnosis and has a relatively high recall rate;The algorithm model also has higher fault recognition accuracy when the signal-to-noise ratio is the same,and the training time for fault recognition is shorter.Therefore,the improved algorithm has significantly improved performance and is of great significance in the fault diagnosis of digital picture book image sensors.
关键词
图像传感器/故障识别算法/故障诊断算法/准确度Key words
image sensor/fault identification algorithm/fault diagnosis algorithm/accuracy引用本文复制引用
基金项目
2022年度陕西乡村基础教育研究课题项目(SXJY202206)
出版年
2024
NETL
NSTL
万方数据