论文部分内容阅读
分析成像系统的工作过程及常用检测算法,针对激光干扰图像的特点,结合背景杂波的量化方法,提出基于目标背景结构相似度与平方和根的图像尺度。该尺度利用结构相似度在成像检测方面以及平方和根尺度在目标局部对比度方面的优点,适用于激光干扰后检测算法性能的评估。激光视场内干扰与视场外干扰的实验显示:激光干扰使得Otsu算法的分割精度受到较大影响;对于相关检测算法,激光干扰会使相关曲面出现较大的波动,增大了虚警概率。研究结果表明:该尺度能够较好地反映激光干扰图像对于CCD成像检测算法性能的影响,初步验证了尺度的有效性。
The working process of the imaging system and the commonly used detection algorithm are analyzed. According to the characteristics of the laser jamming image and the quantification of background clutter, the image scale based on the similarity and square root of the target background structure is proposed. This scale makes use of the advantages of structural similarity in image detection and square root scale in target local contrast, and is suitable for evaluating the performance of post-laser detection algorithm. Experiments with laser field interference and field interference show that the laser interferometry makes the segmentation accuracy of the Otsu algorithm greatly affected. For the correlation detection algorithm, the laser interference causes a large fluctuation in the correlation surface and increases the false alarm probability . The results show that the scale can well reflect the influence of laser interference image on the performance of CCD imaging detection algorithm, and validate the validity of the scale.