Capailities

A complete acquisition software that makes it easy to connect your cameras, adjust the settings, and preview your image. Quick set up the first time, and it will automatically connect to your camera the next time.

Directly view your recorded images and spectrum for quick and easy quality control of your recorded data. Apply reflectance targets to convert new raw images to reflectance, pseudo absorbance and radiance.

Breeze Recorder connects to supported scanner tables and field scanners that synchronises with your camera frame rate to provide seamless data acquisition. View list of supported hardware.

A wide selection of image file formats, point spectral data and other text and tables are supported.  Images files: ENVI (bil, bsq, raw), mat, hdf, hyspex, pam, sc, sac, hips, jpg, png, tiff, and more… Single spectrum / library data: asd, txt, csv, xls, spc, mcf, mat, and more… Other text and data: xls, txt,

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

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