基于Apriori的视觉传感网络关联数据高效识别方法  

Efficient recognition method of associated data in visual sensor network based on Apriori

作　　者：萧展辉 崔焱 李文俊 明哲 杨荣霞 XIAO Zhanhui;CUI Yan;LI Wenjun;MING Zhe;YANG Rongxia(China Southern Power Grid Digital Grid Research Institute Co.,Ltd.,Guangzhou 510663,China;Guangzhou Suinengtong Technology Co.,Ltd.,Guangzhou 510620,China)

机构地区：[1]南方电网数字电网研究院有限公司,广东广州510663 [2]广州穗能通科技有限责任公司,广东广州510620

出　　处：《电子设计工程》2024年第8期130-133,138,共5页Electronic Design Engineering

摘　　要：在视觉传感网络关联数据识别中,由于数据变化的规律难以捕捉,造成识别结果精确度降低,为此,提出了基于Apriori的视觉传感网络关联数据高效识别方法。通过设计视觉传感网络联合监控系统监控视觉传感网络数据变化情况,获取视觉传感网络关联数据。该系统由处理器单元、数据传输单元等四个模块构成。提取图像中的感兴趣区域完成ROI提取,并通过腐蚀处理等方式对视觉传感网络关联数据实施形态学处理。基于Apriori算法实施视觉传感网络关联数据之间关联规则的挖掘,实现视觉传感网络关联数据的识别。布设实验环境后,使用三个数据集测试该方法的识别性能。测试结果表明,所提方法的微平均查全率均高于94.86%,微平均查准率高于93.02%,最高召回率为98.52%,三个项目的测试值均较高,证明了该方法对视觉传感网络关联数据识别比较准确。In the recognition of associated data in visual sensor networks,the accuracy of recognition results is reduced because the law of data change is difficult to capture.Therefore,an efficient recognition method of associated data in visual sensor networks based on Apriori is proposed.The visual sensor network joint monitoring system is designed to monitor the changes of visual sensor network data and obtain the associated data of visual sensor network.The system consists of four modules:processor unit and data transmission unit.The ROI is extracted by extracting the region of interest in the image,and the morphological processing is carried out on the associated data of the visual sensor network by means of corrosion processing.Based on Apriori algorithm,the association rules between the associated data of visual sensor network are mined to realize the recognition of the associated data of visual sensor network.After setting up the experimental environment,three data sets are used to test the recognition performance of this method.The test results show that the micro average recall rate of the proposed method is higher than 94.86%,the micro average precision rate is higher than 93.02%,and the maximum recall rate is98.52%.The test values of the three items are high,which proves that the method is more accurate in identifying the associated data of visual sensor network.

关 键 词：APRIORI算法 FPGA芯片 ROI提取 视觉传感网络 关联数据识别 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]