Spectra Insight™ Software


Main Application Screen: Menu (top), database (left), Image plane of selected bandwidth(s) with cross-sections and histogram (middle), Spectrum at cursor position (upper right), Statistics of entire cube (lower right).
Hyperspectral Cube: The Cube is like a 3D x-ray, where all spectral regions within the cube are viewable by rotating, panning, zooming, and slicing through the cube.
​
​
​
Overview
Spectral Insight acquires and processes spectral data in near-real time from virtual any imaging spectrometer. Spectral data can be imported from 3rd party sources for analysis. Image size/resolution and spectral range/resolution are limited only by the imaging spectrometer.
​
Spectral Insight utilizes fast voxel processing to display thousands of spectral images as a movie or as a rotating 3D volume in real-time. Each planar image or 3D volume are assigned unique color palettes associated with a wavelength or set of wavelengths. For example, wavelengths in the blue region can be given a blue scale pallet ranging from 0-255 while images in the infrared wavelength region are given a grey scale palette ranging from 0-255.
Within the 3D volume, virtual images are stacked and displayable in three different planes xy, xz, and yz and displayed as a movie. There is a full spectrum at each xy position, a set of y-coordinates as a function of x-position and wavelength, and a full set of x-coordinates as a function of y-position and wavelength. A spectrometer with 2048 wavebands scanning across 1000 xy-points provides 2048+1000+1000 (4048) unique images to display and analyze for unique characteristics of each measurement sample. Each image plane is color coded to enhance important aspects of the measurement sample.
​
Each pixel, region, spectral image, and 3D volume can be analyzed for anomalies and/or compared with native or 3rd party spectral databases.
Partial Feature Summary:
​
-
Spectral data acquisition over user selectable areas and resolutions.
-
Selectable wavelength ranges and bandwidths (see spectrometer options).
-
Thousands of spectral bands displayed as images or a movie.
-
Unique color palettes for ultra-violet, visible, and infrared wavelengths.
-
Search for anomalies.
-
Search local or online spectral database for signature matching.
-
Analyze entire images, selectable regions, or individual pixels in seconds.
-
Display volumetric, 3D Surface, 2D Image, cross-sections, and pixel spectral data.
-
Spectral normalization using industry standards such as Spectralon White or any desired spectral data.
-
Import 3rd party spectral data for analysis.
​
New in Version 2.0
​
-
Data Reduction: Three methods including Principle Component Analysis (PCA)
-
Enhanced search with up to 10 classes per second over the entire image
​
New in Version 1.2 !
​
-
Parallel Processing now allows real-time visualization and statistical analysis of any region within a Hyperspectral 3D Cube containing thousands of images, each at a different wavelength.
-
Using parallel processing, the entire Hyperspectral 3D Image can be scanned for a spectral match of any desired substance within one second using an Intel i7 4-Core 8-Thread processor. For the Hyperspectral image in Figure 1A, that’s 60,000 spectral searches/matches in one second. Much higher speeds can be achieved with higher end computers and/or additional processers without any change to the Spectra Insight software.
-
Pseudo RGB color images can now be simulated across any spectral range.
-
Multiple wavelength planes can be analyzed individually or simultaneously.
​
Flexibility
​
-
See our Hyperspectral Spectrometers.
-
See our Gallery for applications and real-world examples.
-
​
-
New features added upon request.
​

Color of cross sections is average of that particular color band. Grey cross section is the near infrared band (780 to 1100 nm).