GUIDE TO SELECTING HYPERSPECTRAL INSTRUMENTS

Similar documents
High Speed Hyperspectral Chemical Imaging

What Makes Push-broom Hyperspectral Imaging Advantageous for Art Applications. Timo Hyvärinen SPECIM, Spectral Imaging Ltd Oulu Finland

SPECIM, SPECTRAL IMAGING LTD.

MEASURING CRUST COLOR WITH HYPERSPECTRAL IMAGING

Hyperspectral goes to UAV and thermal

Chemical Imaging. Whiskbroom Imaging. Staring Imaging. Pushbroom Imaging. Whiskbroom. Staring. Pushbroom

SPECTRAL SCANNER. Recycling

Spectral Analysis of the LUND/DMI Earthshine Telescope and Filters

Near-IR cameras... R&D and Industrial Applications

Applications of Steady-state Multichannel Spectroscopy in the Visible and NIR Spectral Region

Specim. Making spectral imaging possible

Improving the Collection Efficiency of Raman Scattering

Signal-to-Noise Ratio (SNR) discussion

Industrial Applications of Spectral Color Technology

Hyperspectral Systems: Recent Developments and Low Cost Sensors. 56th Photogrammetric Week in Stuttgart, September 11 to September 15, 2017

How does prism technology help to achieve superior color image quality?

Observational Astronomy

sensors ISSN

WHITE PAPER MINIATURIZED HYPERSPECTRAL CAMERA FOR THE INFRARED MOLECULAR FINGERPRINT REGION

DESIGN AND CHARACTERIZATION OF A HYPERSPECTRAL CAMERA FOR LOW LIGHT IMAGING WITH EXAMPLE RESULTS FROM FIELD AND LABORATORY APPLICATIONS

COLOUR INSPECTION, INFRARED AND UV

Miniaturized hyperspectral imaging cameras

WITec Alpha 300R Quick Operation Summary October 2018

APPLICATIONS FOR TELECENTRIC LIGHTING

Hyperspectral Systems: Recent Developments and Low Cost Sensors

Early detection of melanoma using multispectral imaging and artificial intelligence techniques

Spatially Resolved Backscatter Ceilometer

Hyperspectral imaging in Industrial Machine Vision

The chemical camera for your microscope

Imaging with hyperspectral sensors: the right design for your application

Hyperspectral imaging (HSI) goes embedded All rights reserved Max Larin, 1

Optimizing throughput with Machine Vision Lighting. Whitepaper

CMOS BASED HYPERSPECTRAL IMAGING FOR COMPACT / LOW-COST / HIGH-VOLUME IMAGING SPECTROSCOPY. SPIE Baltimore - April 2016

POTENTIAL OF MULTISPECTRAL TECHNIQUES FOR MEASURING COLOR IN THE AUTOMOTIVE SECTOR

Performance Comparison of Spectrometers Featuring On-Axis and Off-Axis Grating Rotation

Speed and Image Brightness uniformity of telecentric lenses

CONFIGURING. Your Spectroscopy System For PEAK PERFORMANCE. A guide to selecting the best Spectrometers, Sources, and Detectors for your application

Hyperspectral Imager for Coastal Ocean (HICO)

Interference metal/dielectric filters integrated on CMOS image sensors SEMICON Europa, 7-8 October 2014

Spectral signatures of surface materials in pig buildings

Hyperspectral Imaging Basics for Forensic Applications

StarBright XLT Optical Coatings

TriVista. Universal Raman Solution

The FTNIR Myths... Misinformation or Truth

ECEN. Spectroscopy. Lab 8. copy. constituents HOMEWORK PR. Figure. 1. Layout of. of the

A novel tunable diode laser using volume holographic gratings

Multi-spectral acoustical imaging

R. J. Jones Optical Sciences OPTI 511L Fall 2017

Order Overlap. A single wavelength constructively interferes in several directions A given direction can receive multiple wavelengths.

Development of four-dimensional imaging spectrometers (4D-IS)

Chapter 36: diffraction

Spark Spectral Sensor Offers Advantages

Optical basics for machine vision systems. Lars Fermum Chief instructor STEMMER IMAGING GmbH

Fast Laser Raman Microscope RAMAN

FT-IR IMAGING THAT'S CLEARLY MEASURABLY AMAZING. Spotlight 400 FT-IR and 400N FT-NIR Imaging Systems

Tunable narrow-band filter for imaging polarimetry

Fast Laser Raman Microscope RAMAN

photolithographic techniques (1). Molybdenum electrodes (50 nm thick) are deposited by

Photonics and Optical Communication

CHAPTER-4 FRUIT QUALITY GRADATION USING SHAPE, SIZE AND DEFECT ATTRIBUTES

LITE /LAB /SCAN /INLINE:

Automated Spectrophotometric Spatial Profiling of Coated Optical Wafers

Supplementary Information for. Surface Waves. Angelo Angelini, Elsie Barakat, Peter Munzert, Luca Boarino, Natascia De Leo,

NIR SPECTROSCOPY Instruments

Quantitative Hyperspectral Imaging Technique for Condition Assessment and Monitoring of Historical Documents

Design Description Document

How to Choose a Machine Vision Camera for Your Application.

Radial Polarization Converter With LC Driver USER MANUAL

Application of a Liquid Crystal Tunable Filter to Near-Infrared Spectral Searches

Photonic-based spectral reflectance sensor for ground-based plant detection and weed discrimination

Advanced Camera and Image Sensor Technology. Steve Kinney Imaging Professional Camera Link Chairman

Supercontinuum based mid-ir imaging

Vixar High Power Array Technology

MEMS Spectroscopy Overview

Design of Efficient Filters for Full-Color Displays Used with Night Vision Devices.

Copyright 2004 Society of Photo Instrumentation Engineers.

Fastest high definition Raman imaging. Fastest Laser Raman Microscope RAMAN

In sum the named factors cause differences for multicolor LEDs visible with the human eye, which can be compensated with color sensors.

MICRO SPECTRAL SCANNER

High-speed Micro-crack Detection of Solar Wafers with Variable Thickness

Applied Machine Vision

Observing a colour and a spectrum of light mixed by a digital projector

Bruise Detection Using NIR Hyperspectral Imaging for Strawberry

UV LED ILLUMINATION STEPPER OFFERS HIGH PERFORMANCE AND LOW COST OF OWNERSHIP

Light Emitting Diode Illuminators for Video Microscopy and Machine Vision Applications

Flat Top, Ultra-Narrow Band Pass Optical Filters Using Plasma Deposited Hard Oxide Coatings

Resampling in hyperspectral cameras as an alternative to correcting keystone in hardware, with focus on benefits for optical design and data quality

Waveguide-based single-pixel up-conversion infrared spectrometer

NIR - SPECTROSCOPY. Sorting technology comparison

Satellite/Aircraft Imaging Systems Imaging Sensors Standard scanner designs Image data formats

Supplementary Materials

Reprint (R37) DLP Products DMD-Based Hyperspectral Imager Makes Surgery Easier

Method for quantifying image quality in push-broom hyperspectral cameras

Spotlight 150 and 200 FT-IR Microscopy Systems

On spatial resolution

BaySpec SuperGamut OEM

06SurfaceQuality.nb Optics James C. Wyant (2012) 1

Introduction to the operating principles of the HyperFine spectrometer

Study guide for Graduate Computer Vision

Color Measurement with the LSS-100P

Transcription:

GUIDE TO SELECTING HYPERSPECTRAL INSTRUMENTS

Safe Non-contact Non-destructive Applicable to many biological, chemical and physical problems Hyperspectral imaging (HSI) is finally gaining the momentum that it deserves as a new vision technology for industrial on-line quality control, inspection and process monitoring. Applications cover wide range of fields from food sorting to printing, recycling and vegetation to name a few. Compared to more familiar black and white or RGB -imaging it provides more detailed chemical, physical and biological information for automation and decision making. Technological power comes from measuring the information above visible wavelength range of light to extract relevant information. Ambitious machine vision companies that want to stay at the technological forefront should not bypass current opportunity to adopt HSI as an important competition asset. Recently, many different HSI technologies have appeared as potential alternatives and it may turn out to be difficult to make reasonable comparison. This short article gives you a basic understanding on a few selection criterias and shows the major differences between the different technologies. Hyperspectral imaging has already shown to be viable solution in many cases, however the selection of right technology for solving any industrial measurement tasks is always a key step to success.

Table of Contents 01. General considerations 02. Illumination requirements 03. Light collection efficiency 04. Summary Specim, Spectral Imaging Oy Ltd. 2017 Guide to selecting hyperspectral instruments ver 1-2017

01 - General considerations When selecting instrument for hyperspectral application there are many functional and non-functional quality attributes to consider. Typically, spectral range, resolution, speed and cost (return of investment) set by application are the first ones to think of. However, one should also consider how selected technology affects the need for illumination, light collection efficiency, purity of spectral data, accuracy and simplicity of data processing. Main technological differences are related to how areal and spectral measurement is carried out in practice. It is not possible to measure two dimensional spatial image and spectrum from each position simultaneously. This dilemma divides current instruments as follows. Push-broom based instruments Push-broom instrument is based on matrix camera and spectrograph to form a line imaging device where one axis of the detector registers spatial position from a line and the other axis spectral information in each spatial position. This assures that resulting spectrum is perfectly co-registered so that all spectral bands are measured simultaneously and from the exactly same sample position. Illumination is needed only for this narrow line and result can be processed immediately after each line measurement. Other HSI technologies Current main competing technologies are tunable filter and variable filter instruments. In tunable filter only one wavelength band is measured from two-dimensional area at each time. In linear variable filter all wavelength bands are measured

simultaneously but each band from different position on the target area. Both suffer from difficulty in getting coregistered spectrum and wasting most of the illumination energy due to operation principle. Co-registration problems lead to difficulties in data processing, unreliable spectral signatures and delayed processing results. Furthermore these technologies require homogenous and stable illumination for much wider two-dimensional area. Push-broom Tunable filter Linear variable filter Data co-registration Good Poor Poor Illumination efficiency Good Poor Poor Spectral resolution Good Medium Poor Spectral purity Good Poor Poor Spatial resolution / image quality Good Good Medium Transmission throughput Good Medium / Poor Medium Table: Basic quality attributes of different HSI technologies while used in industrial on-line applications. Push-broom for industrial on-line applications It is reasonable to conclude that only line imaging push-broom HSI is suitable to all industrial on-line applications where moving and changing targets are measured and spectral coregistration is required. This property is necessary to assure spectral purity and for getting reliable classification results. Similar exact co-registration is practically impossible to achieve with any other methods. Push-broom gives instant measurement result because one does not need to wait for spectral scanning or sample movement to receive all spectral components.

02 - Illumination requirements Hyperspectral imaging needs more light than ordinary machine vision systems. This is fundamental and independent on used technology as light is always divided to narrow wavelength bands and each band is measured separately. Therefore, each pixel receives less light during fixed integration time than compared, for example, to RGB color imaging. However, requirements for illumination are also intrinsically very different within available HSI technologies. Lighting for line imaging Push-broom instrument is line imaging and all wavelengths are collected simultaneously from this line. Therefore, while using push-broom HSI instrument one only needs to illuminate a narrow line in a conveyer belt or sample. Illustration: Example of two line light setup.

In most cases this is accomplished by having two line light sources in 45 angles providing reasonably good and homogenous illumination also with samples having height and surface structure variance. Lighting for area imaging On the other hand tunable filter and linear variable filter devices are area imaging and require illumination within much larger surface area and with homogenous brightness. Generating uniform and stable illumination over an area is much more challenging than over a line. This is especially true as most light sources are point sources and having homogenous illumination means long distance from sample area leading to additional losses. Exposure time and heat load Focusing the light energy along a line will easily provide 30 times higher intensity with the same total source power than illuminating a square area. In other words this means that push-broom HSI can have at least 30 times shorter exposure time and 30 times higher image rate that filter systems. On the other hand, if the speed is not the primary requirement, a push-broom system can work at 30 times lower light power and heat load on the sample. This may become a very essential benefit with biological, medical and food samples.

03 - Light collection efficiency Not all hyperspectral instrument technologies are equal when it comes to how well it utilizes available illumination. Light collection efficiency in a HSI instrument depends on following parameters: Light throughput of optics Polarization characteristics Optical transmission (which includes reflection and absorption losses in optical elements and spectral efficiency of the element which separates the spectral bands) Throughput The throughput is determined by the maximum solid angle by which the optics can collect light from the object (usually related to F number of optics) and transfer it to the detector multiplied by the area of the detector element. As an example here, we are taking a look at a NIR hyperspectral imager with spectral resolution of 10 nm and having 640 x 256 pixels of 24x24 microns each referring to current InGaAs or a MCT detector arrays. Note that the same reasoning applies also to silicon based devices in VNIR range. F-number The maximum light collection angle in imaging optics is determined by its F-number. High quality objective lenses and imaging spectrographs are working with an F-number of F/1.7, resulting in the total collection angle of 34 and corresponding solid angle of 0.28 sr. In an imaging system, the light detection

element is one pixel. Multiplying the solid angle by the pixel area, we get the light throughput value of 16 x10-6 cm2sr for a high performance push-broom hyperspectral imager. In all tuneable filter technologies, the maximum incident angle has to be kept much lower (corresponding to higher F-number). High quality objective lenses and imaging spectrographs are working with an F-number of F/1.7 Throughput with other technologies For the 10 nm resolution in a Liquid Crystal Tuneable Filter (LCTF), the total incident angle has to be limited to ca. 7 corresponding to F/8. LCTF s are currently available in the size of 20 mm in diameter. It is larger than the detector, and thus the optics from the filter to the detector can have a demagnification. This increases the total incident angle on the detector to 10.5 corresponding to a solid angle of 0.025 sr. Multiplying this by the pixel area, we have the light throughput of 2.4 x10-7 cm2sr. This means that LCTF s have nearly two orders of magnitude lower transmission compared to pushbroom imagers. Acousto-optical tuneable filters (AOTF) have the same or even tighter limitation to the incident angle as LCTF s, and their active areas are not larger than 10 mm in diameter in current technology. This results in even lower light throughput. Fabry- Perot type tuneable filters with large areas are available, but we leave them out from this consideration because of their

relatively narrow tuneable spectral range. However, they have same limitation as other tuneable filters due to limited F-number. Linear variable filters have better efficiency compared to tuneable filters mainly due to higher transmission properties and large area. Due to operation principle (deposited Fabry- Perot filter) the input solid angle is limited thus reducing throughput to same values as LCTF. Polarization and spectral transmission The technologies differ also in terms of polarization and spectral transmission, the biggest difference being in light polarization behavior. When push-broom HSI is based on a transmissive spectrograph it is practically independent of polarization and efficiency is only limited by diffraction properties. Tuneable filters light loss All the tuneable filters require linearly polarized incident light in order to work properly, and must have a linear polarizer in the input. It automatically results to 50% light loss with randomly polarized incoming light. Linear variable filter is better in this respect because light does not need to be polarized. Other factors affecting the optical transmission and total efficiency are fairly similar in all the technologies and do not have major contribution to total energy budget.

Summary of collection efficiency Summarizing the analyses done above, we can conclude that compared to push-broom other technologies require from 20, up to 50 times more light or by the same order of magnitude longer exposure times per image to have same signal to noise ratio. This analysis shows considerable difference in the light collection efficiency of HSI technologies. Combining the results from both the light collection efficiency and illumination efficiency, we come to the very significant conclusion that push-broom type HSI is capable of working at 450 to 900 times shorter exposure time and equally higher image rates.

04 - Summary 1. Push-broom technology is the only suitable technology for all on-line application where you have to measure moving and changing targets with spectral co-registration. 2. Hyperspectral imaging needs light, if you want to have easy and cost effective lighting setup and your objects suffer from high heat load then make sure that you select push-broom based instrument. 3. High quality objective lenses and imaging spectrometers are working with an F-number of F/1.7. 4. Aim for high light throughput.

About the Author Esko Herrala is a co-founder of Specim and has been working in the field of hyperspectral imaging since 1992. At the moment he is taking part on sales and marketing activities as a senior expert in the field of machine vision and industrial applications. Follow me on Twitter: @EskoHerrala Specim is one of the recognized leaders in the international hyperspectral imaging industry with a strong market position in a variety of applications. Specim has reached a strong position in the global markets as an instrument, systems and embedded software manufacturer. For more information visit www.specim.fi