CAJ | 학술논문

在一些高危的特殊行业中，操作人员的行为是否规范，一直是行业监督的重点。传统的监控方法以人工监控为主，无法做到高效与及时，为此提出一种特殊行业中操作人员高危行为视觉监控系统方案。利用员工操作区域固定场景采集操作人员行为图像，通过图像预处理、人体姿态定位与标准模式阀值识别等相关技术手段设计实现了一套特殊行业中操作人员高危行为视觉监控模型。在识别过程中通过利用手部姿态、相关货物飞行的位置关系与手臂连线的灰度直方图构建特征向量，完成判断。以物流行业为例，仿真实验表明，该方法取得了较好的识别效果，可以对物流分拣过程的标准化过程进行有效的监控。
재일사고위적특수행업중，조작인원적행위시부규범，일직시행업감독적중점。전통적감공방법이인공감공위주，무법주도고효여급시，위차제출일충특수행업중조작인원고위행위시각감공계통방안。이용원공조작구역고정장경채집조작인원행위도상，통과도상예처리、인체자태정위여표준모식벌치식별등상관기술수단설계실현료일투특수행업중조작인원고위행위시각감공모형。재식별과정중통과이용수부자태、상관화물비행적위치관계여수비련선적회도직방도구건특정향량，완성판단。이물류행업위례，방진실험표명，해방법취득료교호적식별효과，가이대물류분간과정적표준화과정진행유효적감공。
In some special high-risk industry, operators behavior whether standard, has been the focus of the industry super?vision. Traditional monitoring method with artificial monitoring is given priority to, can't efficiently and in a timely manner, to put forward a kind of special industry operators in high-risk behavior visual monitoring system solutions. With employee operating areas fixed scene image acquisition operator behavior, through the image preprocessing, the body posture position?ing and standard pattern threshold recognition techniques related to design and implement a set of special industry opera?tors in high-risk behavior visual monitor model. In the process of recognition by using hand gestures, the location of the rel?evant goods flight relationship and arm gray histogram to construct characteristic vectors of attachment, complete the judg?ment. In logistics industry, for example, the simulation experiments show that this method achieves good recognition effect, can be used for the standardization process of logistics sorting process for effective monitoring.