基于红外虚拟偏振的目标增强方法

Infrared surface target enhancement based on virtual variational polarization

研究了一种基于虚拟变偏振理论的红外面目标细节增强方法,该方法充分挖掘和利用了红外偏振信息的固有特点,利用入射光Stokes矢量和出射光光强之间的描述关系推导出出射光光强和起偏角度的解析关系,并据此虚拟实现了任意起偏角度的出射光光强,然后依据面目标信杂比最大准则,利用粒子群算法迭代实现了最优起偏角度的搜索,最终得到增强后的红外面目标图像。利用实测的长波红外偏振图像数据对算法的可行性和有效性进行了验证,实验结果表明,在灰度对比度、平均梯度、图像熵3种图像质量评价指标下,经此算法增强后的图像质量更高,为后续目标细节识别和攻击点选择等算法的实现奠定了良好的数据基础。

A surface target enhancement method based on the virtual variational polarization theory is pro posed, which makes full use of the inherent eharacteristics of infrared polarization information. By using the de- scription relationship between the stokes vector of the incident light and the intensity of the emitted light, the analytical relationship between the intensity of the emitted light and the polarizing angle is derivated, and thus virtually realizes the intensity of the emitted light with any polarizing angle; then according to the surface target eriterion of the maximum signal-to-noise ratio, the searching of optimal polarizing angle is iteratively realized by using the particle swarm algorithm, and finally get the enhanced infrared surface target image. The feasibility and validity of the algorithm are validated by using the real long wave infrared polarization image. Experimental results show that, the enhanced image using the proposed algorithm has better quality under three image quality evaluation indexes of gray scale contrast, average grads, and image entropy, which lays a good foundation for realizing the following algorithms of target detail recognition and attacking point choosing.