I want to generate the test & train data samples from the hyperspectral image of indian pines & pavia university.
I also want to know a single pixel have multiple values due to bands. Which value is taking for training & testing process.
I am using Matlab as a simulation tool.
http://www.ehu.eus/ccwintco/index.php?title=Hyperspectral_Remote_Sensing_Scenes
In the paper link below, they randomly sample 200 pixels for each class and the training set consists in 5x5 neighborhoods around these pixels. The testing set consists in all remaining pixels.
Article Contextual Deep CNN Based Hyperspectral Classification
A different sampling method.
Pixels selected for training, classification model applied on all the images and 50 pixels per image were randomly selected as ground truth.
Article Ground-level hyperspectral imagery for detecting weeds in wheat fields
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training