Feature detection using electromagnetic tomography and neural networks

作     者：Lau, Jose Nuno Borges, Antonio Rui 

作者机构：Universidade de Aveiro Aveiro Portugal 

出 版 物：《IEE Colloquium (Digest)》 (IEE Colloq Dig)

年 卷 期：1996年第143期

页      面：varpaging页

核心收录：

学科分类：070207[理学-光学] 0808[工学-电气工程] 081203[工学-计算机应用技术] 08[工学] 0835[工学-软件工程] 0701[理学-数学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 0702[理学-物理学] 

主　　题：Computerized tomography 

摘      要：Electromagnetic tomography (EMT) was used to perform feature detection of anomalies of `a priori known objects, employing neural networks (NNs) for data processing. Investigations showed that the NNs have good discrimination properties over features for which they have been trained, performing very well even in a not very precise environment where the object was moved by hand. An estimation approach lead to simpler and faster training. For a similar resolution, the network complexity was also much lower than for the classification approach. The EMT NN combination is a very promising approach for applications requiring not an image but information on the variability of a particular feature of a given object.