PAR-Spec
TM
Full-Field Hyperspectral Imaging Microscopy
An Innovative Approach To Hyperspectral Microscopy
Par- Spec acquires wavelength vs intensity spectra that correlate with CIE HSV ( Hue, Saturation, and Value) color-
space.
CIE coordinates are computed as a function of reflection spectra. Consequently, there is an intrinsic connection
between reflection spectra and HSV coordinates. However, this relationship is one directional converting a HSV back to
reflection spectra is not possible.
Following the development of pseudo-colored, classified reference spectral and HSV libraries, spectral content of an
entire FOV becomes a Hyperspectral Image with no further intervention.
In Summary
PAR-Spec Hyperspectral Imaging offers a comprehensive solution for laboratory-based hyperspectral imaging,
meeting many of the goals of traditional hyperspectral imaging.
PAR-Spec eliminates the need for tedious sample translation. User-selected pseudo-colors correlate with target
objects or areas of interest, utilizing spectral libraries analogous to traditional hyperspectral imaging.
Following the creation of a library an entire FOV is acquired in a single acquisition and analyzed based on
%reflection/emission spectroscopy. Each spectrum is defined by hundreds of contiguous wavelength data points from
400 to 920 nm.
Designed to work with objects as small as 1 micron
2
, PAR-Spec is suitable for a wide range of applications, including
particulates, microplastics, fibers, OLEDs, insects, forensic material, minerals, and vegetation.
The PAR-Spec footprint is only 30 x 30 cm (excluding the lamp) makes it ideal for tight laboratory bench spaces.
For more detailed information download the PAR-Spec pdf
How Traditional Push-broom/linescan Hyperspectral Imaging (TradHSI) Works
1. Spectral Analysis – Application Dependent
•
Method: Based on analytical %reflection or
emission spectroscopy in the visible
wavelength range (400-950-nm)
o
TradHSI: Yes
o
PAR-Spec Yes.
2. Field of view (FOV) Capture - Push-Broom
•
Technique: Sequential "push-broom" or
"linescan" scanning.
o
TradHSI: Yes
o
PARSpec: No
•
Sequential Scanning: The sample is moved on
the microscope stage in a series of steps,
capturing the FOV in "slices" or lines.
o
TradHSI: Yes
o
PAR-Spec: No during routine operation. Yes
to add selected targets to a spectral library,
•
Spatial Resolution Each slice contains spectra
from hundreds of spatially resolved objects
within the FOV.
o
TradHSI: Yes
o
PAR-Spec: Yes
•
Spectral Classification: Self or auto-
classification algorithms provide candidate
spectra for inclusion in a reference spectral
library.
o
TradHSI: Yes
o
PAR-Spec: Yes
•
Costly computer controlled stage. Microscopic
objects require a computer controlled stage
with nanometer accuracy and precision
o
TradHSI: Yes
o
PAR-Spec Yes
3. Hyperspectral software
Regardless Of How You Get There, The Essence Of
Hyperspectral Imaging Is A Two-part Process:
1.
Acquire and interpret detailed spectral information to identify and map objects based on
their unique spectral signatures.
2.
Use pseudo-coloring to represent and interpret this spectral information visually
Primary Functions:
•
Spectroscopic analysis and spectral classification
functions.
•
Correlation/self and auto-classification algorithms.
•
Camera and stage control.
•
Diverse other functionalities
•
Representative target objects in a FOV. The spectra
of all objects in the FOV are self or auto classified
and user selected spectral classes are added to a
spectral library.
o
TradHSI: Yes (ENVI and others) includes
concatenation and stage control
o
PAR-Spec: Yes using LightForm HSI software
4. Reference Spectral Libraries
•
Wavelength vs intensity %reflection /emission/
transmission .reference spectral library
o
TradHSI: Yes
o
PAR-Spec Yes
•
Reference spectra are assigned pseudo colors to
indicate the presence of target objects in the FOV.
o
TradHSI: Yes
o
PAR-Spec Yes
5. Data Visualization - The Hyperspectral Image
•
User-selected pseudo colors represent spectral data
that correlates with target objects. Provides a
visually easy interpretation of a complex FOV..
o
TradHSI: Yes
o
PAR-Spec Yes
•
The presence of a particular pseudo-color indicates
the presence of a specific spectrum that, in turn,
indicates the presence of a target object or area.
•
TradHSI: yes
•
PAR-Spec Yes
Figure 3. Hyperspectral Imaging Microplastics the TradHSI way
3a. Push-broom capture of the spectral data of microplastics
The image shows a FOV containing various microplastic shards. The yellow rectangles represents a single slice
across the FOV that has been translated from the beginning to the end of the FOV.
Light from within the slice passes through to the imaging spectrometer which captures the spectral data. The
microscope translation stage typically requires a 10-nanometer step size.
Spectra are acquired simultaneously as the FOV is scanned sequentially.
Slices are subsequently concatenated to form a hyperspectral image of the entire FOV.
3b. Creating a TradHSI spectral library.
After capturing spectral data, a spectral library is generated made up of unique spectral signatures of each target
object within the FOV. Each signature in the library is assigned a pseudo-color for easy identification.
3c. Identifying target objects:
If the spectrum of a target object matches a spectrum in the library, that object will be assigned the same pseudo-
color as its corresponding library spectrum. For instance, if the green pseudo-color is selected, only objects with
spectra matching the "green" spectrum will be visible in the final hyperspectral image. Other objects will not be
shown.
3d. Hyperspectral image:
The final hyperspectral image isolates objects based on their correlation to specific pseudo-colors in the spectral
library. In this example, the image would show only the objects that correlate with the green pseudo-color.
Figure 4: Hyperspectral Imaging Microplastics with PAR-Spec
4a. Raw Image:
This image was acquired with ColSpec for the highest contrast and shows a single spectral slice manually
located to capture as many target objects as possible. No additional slices are needed if all objects of interest
are within this slice. However, if objects of interest are outside this slice, additional slices will be needed to
capture their spectral information
4b. Spectral Library:
This panel displays the spectral library generated from objects within the slice of the FOV. This library is almost
identical to one that would be obtained using traditional hyperspectral imaging (Trad HSI) methods.
4c. HSV/CIE Color Coordinates:
This step involves converting the library spectra into correlated HSV/CIE color coordinates. The example
highlights a broken red HSV value that correlates with spectral class 2, which is then pseudo-colored green for
visual representation.
4d. Hyperspectral Image:
This panel shows the hyperspectral image of the entire FOV based on the HSV color code associated with
spectral class and is virtually indistinguishable from one generated using a traditional HSI system (Figure 3d).
Each And Every Subsequent Sample Requires An Identical Full-field Scan
Most Importantly Following HSV Library Setup
Each And Every Subsequent Sample Only Requires A Single Acquisition
To Produce A Full-field Hyperspectral Image. No Scanning Needed
How Par-spec Single Acquisition Full-field Hyperspectral Imaging Works
PAR-Spec HSV/CIE Classification Methodology
PAR-Spec
Pros
•
Once an spectral/HSV library has been established hyperspectral imaging future sample is a single acquisition
to cover the entire field.
•
Covers a wide range of applications.
•
The height of a slice through the FOV is almost always less than the available pixels on a modern camera chip.
The HSV acquisition covers the entire chip.
•
Spectral sampling target objects in a limited area of the FOV applies to the entire FOV.
Cons
•
Less suited to applications that cover a limited spectral range. For example, some fluorescence experiments
would be challenging.
•
HSV covers a very wide range of wavelengths, however some applications may present spectra outside the
HSV range.
•
If a FOV is essentially monochromatic PAR-Spec will be ineffective
Traditional Hyperspectral Imaging
Pros
•
Covers a wavelength range that includes wavelengths inside and outside HSV .
•
Covers a very wide range of applications
•
Potential for greater sensitivity in differentiating between very similar spectra.
Cons
•
Requires that every sample be sequentially scanned to cover the entire FOV
•
HSV covers a very wide range of wavelengths, however some applications may present spectra outside the
HSV range.
•
If a FOV is essentially monochromatic PAR-Spec will be ineffective
Par-spec And Tradhsi Compared
PAR-Spec vs Traditional HSI Pros and Cons
Par-spec Hardware
PAR-Spec is a system that includes the PARISS® imaging spectrometer mounted on a research-grade optical zoom microscope (ColSpec). Materials can be illuminated in reflection using either a 150W halogen lamp suitable for spectroscopic applications or a high-intensity (40,000 lumen) LED darkfield ring-light (see Figure 1). Additionally, PAR-Spec can be mounted on a research-grade microscope with dual ports, allowing two cameras to operate simultaneously
1.
Acquire %reflection/emission Spectra Of Background And Target Objects Within A Field Of View
(FOV)
•
Add acquired spectra to a reference spectral library.
•
Take advantage of the possibility to capture the spectra of multiple objects within a single acquisition. (See
Figure 4 here)
2.
Convert Reflection Spectral Library Spectra to CIE-HSV Color Space: (Here)
•
Create a HSV color spectral library in CIE-HSV color space based on reference reflection/emission spectra.
•
The Reference HSV library now correlates with any selected target object.
3.
Hyperspectral Image Analysis:
•
Segment/isolate the acquired image into target objects or conditions based on reference HSV library
spectra. (See Figure 4 here)
•
Count and/or determine the abundance (coverage) of each target object.
4. Single-Snapshot Acquisition for Future FOVs:
•
After creating the HSV color spectral library, only a single snapshot acquisition is needed for future
samples and alternate FOVs.
•
The color camera produces an image comprising a matrix of HSV values. Some or all of these
values will correspond to a reference HSV color spectrum and, consequently, to target objects.
Introduction To How PAR-Spec Hyperspectral Imaging Works
PAR-Spec Eliminates Field-Scanning To Render a
Hyperspectral Image In A Single Acquisition
PAR-Spec
TM
Full-Field Hyperspectral Imaging
Microscopy
An Innovative Approach to Hyperspectral microscopy
For more detailed information download the
PAR-Spec pdf
Regardless Of How You Get There The Essence Of
Hyperspectral Imaging Is A Two Part Process:
1.
Acquire and interpret detailed spectral information to
identify and map objects based on their unique
spectral signatures.
2.
Use pseudo-coloring to represent and interpret this
spectral information visually
Figure 1. PAR-Spec mounted on the ColSpec optical
zoom assembly with brightfield and darkfield
illumination
How PAR-Spec single acquisition full-field hyper-
spectral Imaging works
Figure 2: Hyperspectral Imaging Microplastics with PAR-
Spec
2a. Raw Image: This image shows a single spectral slice
manually located to capture as many target objects as
possible. If all objects of interest are within this slice, no
additional slices are needed. However, if there are
objects of interest outside this slice, additional slices will
be needed to capture their spectral information.
2b. Spectral Library: This panel displays the spectral
library generated from objects within the slice of the FOV.
This library is almost identical to one that would be
obtained using traditional hyperspectral imaging (Trad
HSI) methods.
2c. Spectrum of target object
2d. HSV/CIE Color Coordinates: This step involves
converting the library spectra into correlated HSV/CIE
color coordinates. The example the HSV value that
correlates with the green spectrum is “Class 2” pseudo-
colored green for visual representation.
2e. Hyperspectral Image: This panel shows the
hyperspectral image of the entire FOV based on the HSV
color code associated with spectral class 2. This image is
virtually indistinguishable from one generated using a
traditional HSI system.
Most Importantly Following HSV Library Setup
Each And Every Subsequent Sample Only Requires A
Single Acquisition To Produce A Full-field
Hyperspectral Image. No Scanning Needed
PAR- Spec acquires wavelength vs intensity spectra
that correlate with CIE HSV ( Hue, Saturation, and Value)
color-space.
CIE coordinates are computed as a function of
reflection spectra. Consequently, there is an intrinsic
connection between reflection spectra and HSV
coordinates. However, this relationship is one
directional converting a HSV back to reflection spectra
is not possible.
Following the development of pseudo-colored classified
reference spectral and HSV libraries, spectral content of
an entire FOV becomes a Hyperspectral Image with no
further intervention.
PAR-Spec HSV/CIE Classification Methodology
PAR-Spec and TradHSI Compared
1. Spectral Analysis – Application Dependent
•
Method: Based on analytical %reflection or
emission spectroscopy in the visible
wavelength range (400-950-nm)
o
TradHSI: Yes
o
PAR-Spec Yes.
2. Field of view (FOV) Capture - Push-Broom
•
Technique: Sequential "push-broom" or
"linescan" scanning.
o
TradHSI: Yes
o
PARSpec: No
•
Sequential Scanning: The sample is moved on
the microscope stage in a series of steps,
capturing the FOV in "slices" or lines.
o
TradHSI: Yes
o
PAR-Spec: No during routine operation. Yes
to add selected targets to a spectral library,
•
Spatial Resolution Each slice contains spectra
from hundreds of spatially resolved objects
within the FOV.
o
TradHSI: Yes
o
PAR-Spec: Yes
•
Spectral Classification: Self or auto-
classification algorithms provide candidate
spectra for inclusion in a reference spectral
library.
o
TradHSI: Yes
o
PAR-Spec: Yes
•
Costly computer controlled stage. Microscopic
objects require a computer controlled stage
with nanometer accuracy and precision
o
TradHSI: Yes
o
PAR-Spec Yes
3. Hyperspectral software
Primary Functions:
•
Spectroscopic analysis and spectral classification
functions.
•
Correlation/self and auto-classification
algorithms.
•
Camera and stage control.
•
Diverse other functionalities
•
Representative target objects in a FOV. The
spectra of all objects in the FOV are self or auto
classified and user selected spectral classes are
added to a spectral library.
o
TradHSI: Yes (ENVI and others) includes
concatenation and stage control
o
PAR-Spec: Yes using LightForm HSI software
4. Reference Spectral Libraries
•
Wavelength vs intensity %reflection /emission/
transmission .reference spectral library
o
TradHSI: Yes
o
PAR-Spec Yes
•
Reference spectra are assigned pseudo colors to
indicate the presence of target objects in the FOV.
o
TradHSI: Yes
o
PAR-Spec Yes
5. Data Visualization - The Hyperspectral Image
•
User-selected pseudo colors represent spectral
data that correlates with target objects. Provides
a visually easy interpretation of a complex FOV..
o
TradHSI: Yes
o
PAR-Spec Yes
•
The presence of a particular pseudo-color
indicates the presence of a specific spectrum
that, in turn, indicates the presence of a target
object or area.
•
TradHSI: yes
•
PAR-Spec Yes
PAR-Spec
Pros
•
Once an spectral/HSV library has been established
hyperspectral imaging future sample is a single
acquisition to cover the entire field.
•
Covers a wide range of applications.
•
The height of a slice through the FOV is almost
always less than the available pixels on a modern
camera chip. The HSV acquisition covers the entire
chip.
•
Spectral sampling target objects in a limited area of
the FOV applies to the entire FOV.
Cons
•
Less suited to applications that cover a limited
spectral range. For example, some fluorescence
experiments would be challenging.
•
HSV covers a very wide range of wavelengths,
however some applications may present spectra
outside the HSV range.
•
If a FOV is essentially monochromatic PAR-Spec will
be ineffective
PAR-Spec vs Traditional HSI Pros and Cons
In Summary
PAR-Spec Hyperspectral Imaging offers a comprehensive
solution for laboratory-based hyperspectral imaging,
meeting many of the goals of traditional hyperspectral
imaging.
PAR-Spec eliminates the need for tedious sample
translation. User-selected pseudo-colors correlate with
target objects or areas of interest, utilizing spectral
libraries analogous to traditional hyperspectral imaging.
Following the creation of a library an entire FOV is
acquired in a single acquisition and analyzed based on
%reflection/emission spectroscopy. Each spectrum is
defined by hundreds of contiguous wavelength data
points from 400 to 920 nm.
Designed to work with objects as small as 1 micron
2
,
PAR-Spec is suitable for a wide range of applications,
including particulates, microplastics, fibers, OLEDs,
insects, forensic material, minerals, and vegetation.
The PAR-Spec footprint is only 30 x 30 cm (excluding the
lamp) makes it ideal for tight laboratory bench spaces.
Traditional Hyperspectral Imaging
Pros
•
Covers a wavelength range that includes wavelengths
inside and outside HSV .
•
Covers a very wide range of applications
•
Potential for greater sensitivity in differentiating
between very similar spectra.
Cons
•
Requires that every sample be sequentially scanned
to cover the entire FOV
•
HSV covers a very wide range of wavelengths,
however some applications may present spectra
outside the HSV range.
•
If a FOV is essentially monochromatic PAR-Spec will
be ineffective
PAR-Spec Hardware
PAR-Spec is a system that includes the PARISS® imaging spectrometer mounted on a research-grade optical zoom microscope (ColSpec). Materials can be illuminated in reflection using either a 150W halogen lamp suitable for spectroscopic applications or a high- intensity (40,000 lumen) LED darkfield ring-light (see Figure 1). Additionally, PAR-Spec can be mounted on a research- grade microscope with dual ports, allowing two cameras to operate simultaneously
1.
Acquire %reflection/emission Spectra Of
Background And Target Objects Within A Field Of
View (FOV)
•
Add acquired spectra to a reference spectral library.
•
Take advantage of the possibility to capture the
spectra of multiple objects within a single
acquisition. (See Figure 4 here)
2.
Convert Reflection Spectral Library Spectra to
CIE-HSV Color Space: (Here)
•
Create a HSV color spectral library in CIE-HSV
color space based on reference reflection/emission
spectra.
•
The Reference HSV library now correlates with any
selected target object.
3.
Hyperspectral Image Analysis:
•
Segment/isolate the acquired image into target
objects or conditions based on reference HSV
library spectra. (See Figure 4 here)
•
Count and/or determine the abundance (coverage)
of each target object.
4. Single-Snapshot Acquisition for Future
FOVs:
•
After creating the HSV color spectral library, only
a single snapshot acquisition is needed for future
samples and alternate FOVs.
•
The color camera produces an image
comprising a matrix of HSV values. Some or
all of these values will correspond to a
reference HSV color spectrum and,
consequently, to target objects.
