摘要
研究基于图像的精确导弹制导问题,在发生雷达干扰的情况下,采集的SAR图像因为受到干扰,其输出功率图像中不同区域功率值的绝对大小及它们之间的相对关系被破坏,造成待匹配特征发生较大程度突变;传统的基于图像的制导匹配技术多是基于两种图像间的特征一致性进行对比完成匹配,一旦这种突变造成特征不一致性过大,就会形成漏匹配,造成匹配结果不准;提出了一种基于自然语义特征转换的制导图像匹配算法,把提取到的制导区域特征的多个标准量通过自然语义模型,转换成固定的自然语义时域差异特征,运用更为客观的特征语义系数衡量制导区域特征,完成制导特征的精确匹配;实验证明,利用这种算法提高了制导匹配的准确率。
Research based on image accurate missile guidance problem. Happened in radar interference, the acquisition of the SAR image because interference, the output power of image in different areas of the values of the absolute size and power of the relative relations between them are destroyed, cause for matching feature in great degree mutations. The traditional based on image guidance matching technology are based on the characteristics of two kinds of image contrast between consistency complete matching, once the mutations caused by excessive characteristics of consistency, it would form a leak matching, cause matching result is accurate. Put forward based on the semantic feature conversion of natural guidance image matching algorithm, the face of the features to extract more than quantity standard through natural se-mantic models, converted into a fixed natural semantic time domain difference characteristics, use more objective characteristics of semantic coefficient measure face feature, complete guidance characteristics of the precise match. The experiment shows that the application of this al-gorithm enhances the accuracy of guidance matching.
出处
《计算机测量与控制》
北大核心
2013年第4期990-992,共3页
Computer Measurement &Control
关键词
雷达匹配
自然语义
像素配准
radar matching
natural semantic
pixel registration
