A hyperspectral remote sensing image classification online tool.

Server:

• Tornado
• scikit-learn
• numpy
• scipy
• statsmodels

Webpage:

• Bootstrap
• Echarts
• JQuery

in main folder :

python multi_classifier.py

defalut port : 8989

• /
  After inputing image data and selecting classification method (multinomial logistic regression[MLR] or support vector machine[SVM]), you could get classification result reported by text and graphs.And you could select the parameters and the indices of the feature you focus on.

• /stat_feature
  Getting original data statistical feature, such as average, variance.

• /rbf_gamma
  Trying to find the best gamma of RBF kernel funciton.

• /soft_margin
  Trying to find the best soft margin of SVM algorithm.

• /forward_stepwise
  Selecting feature indices by forward stepwise.

• /knn
  Optimizing classification result by k-nearest neighbors.

• /format
  Data format.

It's in test folder

• all_data for /
• whole_data for /stat_feature
• cv_data for /rbf_gamma and /soft_margin
• fs_data for /forward_stepwise
• result for /knn