目标区域局部特征和局部图像质量相结合的激光干扰效果评估

Laser dazzling effect assessment based on local features and image quality in target region

为客观描述光电成像系统激光干扰效果,提出目标区域局部特征和图像质量相结合的干扰效果评估算法。图像的局部特征用特征点描述,图像质量用结构相似度指数描述。该算法利用特征点匹配算法确定场景图像中的目标区域,计算干扰前后目标区域内特征点数量的变化、干扰后目标区域内未饱和面积所占的比重以及目标区域内图像的结构相似度指数,并将上述各参数相乘得到最终的评价指标。利用该方法对典型激光干扰图像进行评估,结果表明:在不同入射功率和不同光斑位置情况下,所提评估指标比单独使用结构相似度的评估指标具有更大的取值范围和更明显的非线性变化特征。这说明:该指标能够反映光电成像系统激光干扰过程的丰富细节,更加适合激光干扰效果的评估。

A laser dazzling effect assessment of optoelectronic systems based on local features and image quality in the target region was proposed. The feature points were used to describe the local feature of the image. The structural similarity index was computed to reveal the image quality. The assessment utilized the point feature matching algorithm to locate the target region in the scene image. Then the ratio of the number of feature points and the unsaturated area and also the structure similarity index in the target region between the original image and dazzled image were computed. The final assessment index was achieved by the product of these three factors above. The implementation of the assessment to a typical dazzled image suggested that this assessment can provide a more broad range of index value and more obvious nonlinear variation than that obtained by only the structural similarity index in the target region,which indicated that this assessment method suggested in this paper has the ability to show more detail of laser dazzling process and more suitable for the assessment to lase-dazzling effect to image system.