DrugChecking:一种多颜色空间下毒驾检测试纸识别方法  被引量:1

DrugChecking:Identification of Saliva Test Strip for the Drug Driving Detection in Multi-color Space

在线阅读下载全文

作  者:魏婷婷 林楠[1] 曹仰杰[1] 魏君飞 杨聪 WEI Ting-ting;LIN Nan;CAO Yang-jie;WEI Jun-fei;YANG Cong(School of Software,Zhengzhou University,Zhengzhou 450002,China;Hanwei Internet of Things Research Institute,Zhengzhou University,Zhengzhou 450002,China)

机构地区:[1]郑州大学软件学院,郑州450002 [2]郑州大学汉威物联网研究院,郑州450002

出  处:《小型微型计算机系统》2021年第1期147-153,共7页Journal of Chinese Computer Systems

基  金:国家自然科学基金项目(61972092)资助。

摘  要:近年来,由"毒驾"所引发的重大交通事故数量不断增长,现有的毒品检测技术由于时效性、便捷性等原因,难以适用于常规道路毒驾稽查,不利于道路现场检测.对此,本文针对毒驾唾检试纸图像的特点,设计并实现了一种高效的毒品检测方法——DrugChecking.该方法首先对唾液检测试纸进行边缘检测,以提取试纸检测区域;其次,针对毒驾唾检试纸的弱边缘问题,DrugChecking在多颜色空间中采用Hough变换提取试纸条区域;然后,使用主成分分析对试纸条区域进行降维处理;最后,采用支持向量机对降维后的数据进行分类.本文方法已在现场采集的毒驾唾检试纸图像上进行了验证,实验结果表明:DrugChecking的识别准确率达到98.04%,能够有效识别毒驾唾检试纸中毒品类别.In recent years,the number of serious traffic accidents caused by'drug-driving'has been increasing.However,current drug detection technologies are difficult to apply to regular road drug inspection due to the limited timeliness and convenience,which is not conductive to the road on-site detection.To solve the problem,an effective method was designed and implemented according to the characteristics of the drug driving test paper image,which is named as DrugChecking.The proposed method firstly detects the edge of saliva test paper to extract test paper area.To analyze multi-edge test paper,DrugChecking uses Hough transform to extract strip area in multi-color space,and then uses principal component analysis(PCA)to reduce the dimension of strip area,and finally uses support vector mechine(SVM)to classify the reduced dimension data.We have verified the proposed method on drug driving test paper image collected on site.Experimental results show that recognition accuracy rate of DrugChecking has reached 98.04%,showing that DrugChecking can effectively identify the drug category in the drug driving test paper.

关 键 词:毒驾 机器学习 多颜色空间 主成分分析 弱边缘 

分 类 号:TP391[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象