一种抗强脉冲噪声干扰的火场人员视觉定位算法

Fire Personnel Visual Localization Algorithm of Anti Strong Impulse Noise Interference

研究火场人员准确定位问题。火场的情况复杂,烟雾和外界环境的干扰使得采集的火场图像包含大量的非连续性脉冲噪声,极易形成强脉冲干扰环境。传统的图像定位算法在强干扰的环境下,很难保留图像的边缘和细节,对非线性噪声的过滤有着明显的缺陷,噪声干扰使得人员图像边缘出现缺失,定位存在较大偏差。为提高定位准确性,提出了一种抗强脉冲噪声干扰的火场人员视觉定位算法。利用像素插值方法能够计算火场图像中的亚像素位移,从而得到火场图像像素形变程度。利用视觉定位方法,计算火场人员的空间位置参数,从而确定火场人员在标准坐标系统中的实际坐标。实验结果表明,提出的算法能够有效提高火场人员定位的准确性,有利于快速救出火场中的被困人员。

Research the accurate localization of fire personnel. The paper proposed a fire personnel visual localization algorithm of strong resistance of impulse noise. Using pixel interpolation method can calculate the fire image of the sub-pixel displacement, thus obtains the fire image pixel deformation degree. Visual positioning method was used to calculate fire personnel space localization parameter, so that the coordinates of fire personnel in the standard coordinates system were determined. The experimental results show that the algorithm can effectively improve the positioning accuracy of fire personnel.