MUSIC algorithms for rebar detection
MUSIC algorithms for rebar detection
复制标题
用于钢筋检测的 MUSIC 算法
DOI：
10.1088/1742-2132/10/6/064006
复制
发表时间：
2013
影响因子：
1.4
通讯作者：
A. Dell’Aversano
中科院分区：
文献类型：
--
作者：
R. Solimene;G. Leone;A. Dell’Aversano
MUSIC(多信号分类)算法用于检测和定位未知数量的散射体，这些散射体相对于波长而言尺寸较小。待检测目标的集合由强散射体和弱散射体组成。这代表了一个对检测目的具有挑战性的散射环境，因为强散射体往往掩盖弱散射体。因此，较弱散射目标的检测并不总是有保证的，并且当破坏数据的噪声水平相对较高时可能完全受损。为了克服这一缺点，本文从应用两阶段MUSIC算法的思想出发，提出了一种新的技术。在第一阶段，检测到强散射体。然后，在仅聚焦于弱散射体的第二阶段中使用关于它们的数量和位置的信息。强调了适当的散射模型的作用，以显著提高在现实场景中的检测性能。
The MUSIC (MUltiple SIgnal Classification) algorithm is employed to detect and localize an unknown number of scattering objects which are small in size as compared to the wavelength. The ensemble of objects to be detected consists of both strong and weak scatterers. This represents a scattering environment challenging for detection purposes as strong scatterers tend to mask the weak ones. Consequently, the detection of more weakly scattering objects is not always guaranteed and can be completely impaired when the noise corrupting data is of a relatively high level. To overcome this drawback, here a new technique is proposed, starting from the idea of applying a two-stage MUSIC algorithm. In the first stage strong scatterers are detected. Then, information concerning their number and location is employed in the second stage focusing only on the weak scatterers. The role of an adequate scattering model is emphasized to improve drastically detection performance in realistic scenarios.