大面积火灾现场的人员视觉识别定位方法仿真  

Simulation of the visual identification and location method for the personnel in large area fire scene

作　　者：刘莹[1] 祝毅鸣[1] 

机构地区：[1]郑州大学西亚斯国际学院,河南新郑451100

出　　处：《计算机仿真》2016年第6期302-305,共4页Computer Simulation

摘　　要：在大面积火场区域进行视觉优化定位中,由于无法预先知道区域大小,传统的定位方法只能随机选取单独区域进行定位识别,区域目标选取存在较大盲目性,一旦选取区域不当,定位过程会耗时较长,缺陷明显,提出基于改进Gabor小波算法的大面积火灾现场的人员视觉识别定位方法。上述方法先融合于色度方差原理对大面积火灾现场的人员目标区域进行定位,利用Gabor小波对火灾的烟雾纹理和面积边缘特征进行表述,组建火灾烟雾变化能量模型和方向角分布模型,提取火灾烟雾变化中人员的动态特征,通过对大面积火灾现场的人员进行动态特征识别,完成了对大面积火灾现场的人员视觉识别定位。仿真证明,改进Gabor小波算法的大面积火灾现场的人员视觉识别定位方法识别的准确率高,定位精确度高。Visual optimized location in a large area of the fire scene is studied. Because it is unable to know the size of the area in advance, the traditional location method can only randomly choose a separate area for location and identification, there is a large blind in area target selection. Once the selected area is not properly, the location process will take a long time, and the defect is obvious. A visual identification and location method for the personnel in large area fire scene based on the improved Gabor wavelet algorithm is proposed. Firstly, the method is integrated with the principle of color variance, and the personnel target area is located in the large area fire scene. The Gabor wavelet is used to describe the texture and area edge feature of the fire smoke. The energy model and direction angle distribution model of the change of fire smoke are built, and the dynamic features of personnel in the change of fire smoke are extracted. Through the identification of the dynamic features, the visual identification and location of per- sonnel in the large area fire scene is accomplished. The simulation results show that the proposed method has high precision rate and location accuracy.

关 键 词：人员动态特征 特征提取 小波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]