INP=(IN1,IN2,...,STAT) or INP=(IMSS,STAT) Input data sets, where INn are single band inputs and IMSS is the input in MSS format. The last data set MUST BE the output data set from STATS.
OUT=OUT Output data set containing classification map
SIZE=(SL,SS,NL,NS) where SL, SS, NL, NS are the starting line, starting sample, number of line, number of samples in the output file. It is a standard Vicar size field.
Starting line of image
Starting sample of image
Number of lines in image
Number of samples in image
MSS=N Denotes that the input data set is in MSS format and contains N spectral bands. (Default is standard format)
USE=(I1,I2,...) Used in conjunction with MSS; denotes which bands are to be used as input to the classifier and their order of use. (Default is to use all bands specified by MSS.)
BAND=(I1,I2,...) Denotes that statistics data set contains statistics of the specified bands in the given order. (Default is the same order as input data)
SIGMA=(R1,R2,...) Denotes that R1 standard deviations are used for the decision boundaries of the first band used, all classes; R2 for the second, all classes; etc. The last value specified is used for all remaining bands. (default=1.0)
CSIGMA=(I1,J1,X1,I2,J2,X2...) Denotes that X1 standard deviations are to be used for class I1 and band J1. Jn = -1 denotes that all bands for class I1 will use X1. The boundaries specified by this parameter set override those specified by SIGMA. J is the Jth band used, not necessarily band J if MSS and USE are specified.
('DONT)
Denotes that the Bayesian algorithm is not involved to resolve ambiguity.
('CHECK)
Denotes that multivariate confidence boundary is checked by the Bayesian
algorithm.
PRIOR=(I1,I2,I3,I4,...) Denotes that band I1 will contain not multispectral values, but index values indicating which set or sets of prior probabilities are to be used in the Bayesian decision rule. I2 indicates the maximum number of sets of prior probabilities associated with band I1. Thus the coding PRIOR=(9,3,10,5) would indicate that bands 9 and 10 contain indices to probability sets, with the values in band 9 ranging from0-3, and the values in band 10 ranging from 0-5.
PROB=([I1],[J1],X1,[I2],[J2],X2,...) Denotes that prior probabilities X1-XN, where N is the number of classes, are to be used in the Bayesian decision rule. If the PRIOR parameter has specified one channel as an index to probability sets, then I1, I2,... are required and indicate the subscript associated with the set. If PRIOR has specified two channels, the I1,J1,...,I2,J2 are required. The J index specifies the band, and the I index specifies the class, to which the set of probabilities applies. Probability values of zero are permitted.
MEAN=(I1,J1,R1,I2,J2,R2,...) Denotes that the STATS mean for Class I1 and band J1 is to be replaced by the value R1. J refers to the actual band number, not the Jth band used. (This is different from the interpretation of band for CSIGMA and SIGMA.)