INTERNAL: ENVI Plugin for See5 Decision Tree Automatic Classification (SDTC)

See5 is a effective data mining tool which is used to generate a decision tree developed by Ross Quinlan based on entropy algorithm. The decision trees generated by See5 can be used for classification, and for this reason, See5 is often referred to as a statistical classifier, and it has been widely used in over 50 contunries . In the field of classification or landcover/landuse mapping based on remote sensing imagery, decision tree classification method is a accurate and effective way to get a classification map regionally and globally. As we know, It has been used in the NLCD(National Land Cover Database in United States) project operated by USGS and MODIS global landcover product(MOD12) operated by Bonston University. However, See5 require data files with specific pattern and ENVI can not read the decision tree generated by See5 directly to classify imagery.
This plugin is written for enabling See5 to be more conveniently and automatically for ENVI users. Its fuctions are as follows:
(1) According to the text file converted from ROI file in ENVI, it can generate three sepcific files required by See5 software, at the same time, you can set the propotion of training data in the total number of sample dataset,and remaining data are used to test classification rules.
(2) it can read the file named '***.out' generated by See5, and execute that rules automatically within ENVI to get the final classification.

More details about extension usage can be found in the attached RAR file.it should be used in ENVI4.8+ version. See5 is available for Windows 2000/Xp/Vista/7.