改进的判别函数对遥感光谱的识别

Identification of Remote Spectra with an Improved Synthetic Discrimination Function

用最小均方误差方法改进光谱识别的判别函数，适合于从多种复合光谱中识别出特定成份光谱，并具有较高的识别率。最后给出对化学模拟剂DMMP光谱的识别结果。

An improved synthetic discriminant function by least-mean-square-error method for spectral identification is proposed, which is suitable for discriminating one specific class from all other classes. It has high discrimination results. Finally, an application to chemical simulant DMMP spectral identification is given.