In order to effectively classify the selection and extraction of feature quantities, we hope that the distance between similar samples should be as small as possible, while the distance between non-similar samples should be as large as possible.
For SAR image target classification, a feature extraction algorithm based on intra-class distance criterion is selected. Assuming that there are c types, let Jw denote the total average squared distance within the class, ie
While judging the size of J, J is the required combination of features.
Research on Target Classification of Synthetic Aperture Radar Images
Nantong Boxin Electronic Technology Co., Ltd. , https://www.bosencontrols.com