论文部分内容阅读
为准确识别在固定位置上执勤人员的位置和姿势,预防其在执勤过程中出现离岗、蹲坐、倚靠、遇袭等异常情况发生,有效维护执勤人员形象,保证执勤人员人身安全,提出了一种基于Kinect体感设备新型的固定位置执勤人员监控系统。该模型基于普通PC平台,利用Kinect获取监控场景中目标人体的骨骼数据和深度数据,进而获取目标人体的三维空间数据;利用线性支持向量机分类器,实现了执勤人员位置和姿势等参数精确判断,对非法闯入警戒区事件及时发出报警信号。经过测试和实际应用,该模型对环境噪声、人物着装、运动速度等具有很强的鲁棒性,识别算法效率高,具有实时性、高可靠性,具有广泛的社会经济效益。
In order to accurately identify the position and posture of the duty officer in a fixed position and to prevent the occurrence of abnormalities such as separation, squatting, leaning against and attacking during the duty, effectively maintain the image of the duty officer and ensure the personal safety of the duty officer, A new type of fixed position attendants monitoring system based on Kinect somatosensory equipment. The model is based on a common PC platform and uses the Kinect to acquire the skeleton data and depth data of the target human body in the surveillance scene so as to obtain the three-dimensional spatial data of the target human body. The linear support vector machine classifier is used to accurately determine the position and posture of the duty officer and other parameters , Issued a warning signal in time to the incident of breaking into the warning zone illegally. After testing and practical application, this model has strong robustness to environmental noise, character dress, movement speed and so on. The recognition algorithm has high efficiency, real-time and high reliability, and has a wide range of social and economic benefits.