| 摘要: |
| 传统单一特征检测方法的检测性能较差，通过多特征联合检测方法可以有效提高检测性能。采用多特征联合检测方法在提高性能之外，也会造成计算量增加以及信息冗余。对此提出了一种基于线性判别分析的海上目标检测方法，将单一特征映射到二维特征空间中，形成两组特征组合，RDPH-RVE特征组合和RPH-TEM特征组合，并在二维特征组合基础上进行降维处理。通过将单一特征映射到二维空间中，减少了海杂波和目标之间的重叠区域，再通过线性判别分析方法，将雷达回波数据在区分性更好的方向进行投影，在保留信息的同时减少了计算量。 |
| 关键词: 特征提取 小目标检测 海杂波 多特征联合 |
| DOI： |
| 分类号:TN951;TN957.51 |
| 基金项目:国家自然科学基金项目（面上项目，重点项目，重大项目） |
|
| Detection algorithm of small target on the Sea Surface based on linear discriminant analysis |
|
|
|
|
| Abstract: |
| The traditional single feature detection method has poor detection performance, and the detection performance can be effectively improved by multi-feature joint detection method. However, the combination of multi-features will not only improve the performance, but also cause the problems of increased calculation and information redundancy. In this paper, a detection method of floating small targets based on linear discriminant analysis is proposed. A single feature is mapped into a two-dimensional feature space to form two groups of feature combinations, namely RDPH-RVE feature combination and RPH-TEM feature combination. The dimension is reduced based on the two-dimensional feature combination. By mapping a single feature into a two-dimensional space, the overlapping area between sea clutter and target is reduced. Then, by means of linear discriminant analysis, the radar echo data is projected in a more discriminating direction, which reduces the calculation amount while retaining the information. |
| Key words: feature extraction small target detection sea clutter Multi-feature combination |