| 摘要 |
<P>PROBLEM TO BE SOLVED: To reduce the dimension of hyper spectrum image (HSI) having a plurality of space pixels (PX). <P>SOLUTION: Calculation of a set (A) of factors related to each PX of HSI data, a set (B) of basic vectors (BV) used for generating the set of factors, acceptance of the maximum error value (C), an error (D) of first set for each PX related to the B, and an error (E) of a further set for each PX related to the subset of B, calculation of %(F) of the number of space pixels having an error exceeding the maximum error value, and calculation of a plurality of reduction factors (G) related to the D and E (G is calculated based on both the F, and the number of spectrum dimensions related to the HSI data) include selection of the maximum reduction factor from the F, and the optimum size of a set of the BV related to them or a subset of the BV. <P>COPYRIGHT: (C)2013,JPO&INPIT |
