CAJ | 학술논문

施工区中部分作业人员安全意识缺乏,不佩戴安全帽作业情况时有发生. 采用机器视觉的相关方法对安全帽自动识别技术进行深入研究,在图像预处理基础上,首先利用混合高斯模型进行前景检测,然后通过对连通域的处理判断其是否属于人体,最后在定位到的人体头部区域实现安全帽的自动识别. 施工区实地实验结果表明,该方法准确率高、环境适应性强,适于在施工区推广应用.
시공구중부분작업인원안전의식결핍,불패대안전모작업정황시유발생. 채용궤기시각적상관방법대안전모자동식별기술진행심입연구,재도상예처리기출상,수선이용혼합고사모형진행전경검측,연후통과대련통역적처리판단기시부속우인체,최후재정위도적인체두부구역실현안전모적자동식별. 시공구실지실험결과표명,해방법준학솔고、배경괄응성강,괄우재시공구추엄응용.
Due to the lack of safety awareness of workers in the construction area, the workers often do not wear the safety helmet.In view of this situation, it uses the method of machine vision to develop the automatic identifi-cation technology of safety helmet.Based on image preprocessing, it mixtures Gaussian model to detect the fore-ground, then determines the connected domain to judge whether it belongs to the human body or not.Finally, it locates the head region of human and realizes the automatic recognition of safety helmet.The result of construc-tion field experiments shows that the method is of high accuracy and good environment adaptability, and is suit-able for application in the construction area.