HSVap: Hyperspectral Visualization and Processing

HSVap supports the use of advanced methods of Machine Learning for analysis of hyperspectral images by combining

HSVap is an Open Source project which is implemented in Java. A beta release will be published in January 2015.

Key Features

Visual Data Exploration

HSVap Exploration

Visualization of hyperspectral images: RGB, single features and classification results

Easy and intuitive generation of labelled training datasets

Data Mining with cluster algorithms

Classification and Active Learning

HSVap Classification

Linear and non-linear classification of hyperspectral data with Support Vector Machines

Automatic parameter optimization

Active Learning for minimizing time investment for training data generation

Visualization of classification results

Feature Selection

HSVap Feature Selection

Calculation of feature relevance with ReliefF, InformationGain or SVM Weighting

Statistical data analysis of correlation coefficients, standard deviation and mean of classes

Database for Vegetation Indices included

Funded by:
Logo bmbf