Analyze

Explore your data with powerful tools designed to quickly detect patterns and anomalies. Breeze offers an interactive experience, making data analysis intuitive and efficient for deeper insights.

PCA (Principal Component Analysis) in Breeze reduces hyperspectral data complexity by transforming correlated spectral variables into principal components. This helps highlight key patterns for easier visualization and data selection. The PCA descriptor simplifies analysis across multiple images and provides insights into spectral contributions, improving data interpretation and understanding of the chemistry of your materials.

Pixel Explore is a feature in Prediktera’s Breeze software that allows users to examine hyperspectral images at the pixel level, providing detailed spectral information for each pixel and for selected areas.  This tool enables precise selection and analysis of regions of interest within an image, facilitating tasks such as manual segmentation and end-member creation. The

Large built-in tool box for advanced analysis and modelling of your data combined with a user-friendly interface that makes it easy and fast to analyse your data.

Perform image analysis on pixel and object level where you have discrete data (i.e. classes A, B or C).

Perform image analysis on pixel and object level where you have continuous variables (e.g. %, mg/l etc) using available methods

Large selection of models that can be manually selected and auto function that proposes the best model based on cross validation. Available models: Neural network Decision tree Support vector machine Random forest Logistic regression Maximum entropy Poisson regression Linear Regression

Chemometrics models are powerful and fast tools for analysis, diagnostics and visualisation of your data. Available models: PLS PLS-DA Hierarchical PLS-DA,  SIMCA,  PCA,  K-means Gaussian mixture mode