Multi spectral imaging sensors

Course: BREE504 (Instrumentation and Control) Presentation
By: Lanrewaju Adetunji (260606673)
Date: Thursday, October 29th, 2015

 Definition of Multi-Spectral Imaging (incl HSI);
 The Spectrograph;
 Detectors or Image Sensor (incl Types and Noise);
 Resolution, Precision, and Accuracy of MSI systems;
 Application of MSI systems within the Bioresource Engineering domain
 Conclusion
Thursday, 05 November 2015Spectral Image Sensor | Adetunji 2

 Spectoscopy + Conventional Imaging
= Spatial + Spectral Information
 Hyperspectral Imaging > MSI
 EM radiation classified–radio wave,
MW, IR, visible light, UV, X-rays,
gamma rays–according to wave length
 Target illuminated by tungsten-
halogen or LED source or natural
lighting (Sun)
 Spectrograph scatters and measures
reflectance, absorbance, or
fluorescence
 Two-dimensional spatial image (x×y) ×
wavelength (spectral; λ) dimension =
hypercube
Fig 1: The Concept of Imaging Spectroscopy Source:
Shaw & Burke, 2003

Fig 2: Techniques for acquiring hypercube
dataset

The Slit
-Controls
amount and
angle of entry
light
-10, 25, 50,
100, and
200μm
Collimator
-Essentially a
concave mirror
The diffraction grating
-determines λ
range and
optical
resolution of the
system
-2 types: ruled
and
holographic
gratings
Camera lens
-Refocuses
dispersed light
onto the
detector pixels
Detector
-Made up of
several pixels
-Two types:
CCD and
CMOS
The
Spectrograph
Scheme 1: Components and flow of radiation in a spectrograph

 Converts captures EM radiation (light) into
electrical signal (charges);
 Charges iteratively converted into voltage
(AMPLIFIER)
 Additional circuitry converts voltage into
digital information (A/D Conv.)
 Depending on SC material used (Si or
InGaAs),
λ 𝑚𝑎𝑥 =
ℎ × 𝑐
𝐸𝑔𝑎𝑝
where: λ, upper wavelength limit; h,
Planck’s constant; c, speed of
light; Egap, bandgap energy of the SC
Fig 3: Operation of a CCD Image Sensor (Source:
Wikipedia)

 Broadly they can be classified into
two, namely: Analog and Digital
 Digital image sensors are
semiconductor-based:
 Charge-coupled detector (CCD)
 Active pixel sensors e.g. including
Complementary Metal-Oxide
Semiconductor (CMOS)
 Common SC materials:
 Si (1.1eV)
 HgCdTe
 InGaAs
CCD CMOS
DOMAIN
Readout electronic
incorporation
No Yes
Radiation tolerance Lower Higher
Frame shift smear Susceptible Not
Susceptible
Power consumption Higher Lower
System size Larger Smaller
Cost Higher Lower
Rolling-shutter effect Not susceptible Susceptible
Table 1: Comparison of two common SC-based image
sensors used in Multi-spectral Imaging

 Resolution reported as spectral (or optical)
resolution: Determined by:
 the slit; the diffraction grating; the detector.
 Spatial resolution: Typical array sizes in
pixels (or pixel resolution) in many imaging
sensors vary from 640×480 to 2,048×1,536
pixels. For reference, human vision is >100
million pixels
δλ =
𝑅𝐹 × ∆λ × 𝑊𝑠
𝑛 × 𝑊𝑝
where: δλ, spectral resolution of
spectrometer; RF, resolution factor; Δλ,
spectral range; Ws, slit width; Wp, pixel
width; n, number of pixel.
E.g., for a spectrometer with a 25μm slit, a
14μm 2048 pixel detector and a λ range from
350 – 1050nm, calculated resolution will be
1,53nm.
Precision and resolution depend on sampling
resolution

Fig 4: CCD sensor (Source: Wikipedia)
Fig 5: CMOS sensor (Source: Wikipedia)

 Higher 1/f noise in CMOS
 Noise increases with increasing
illumination, but SNR also increases
with illumination
Noise Types/Sources in CMOS:
 Intrinsic sources (Shot, thermal, and
flicker noises);
 Reset noise;
 Pixel noise;
 Readout noise;
 Quantization noise;
 Noise from power supply fluctuation;
 Noise from environmental influences

Fig 6: CCD sensor showing array detectors (pixels) and amplifier

MULTI-SPECTRAL IMAGING:
 Remote sensing;
 Precision agriculture;
 Food quality/inspection: defect
and contaminant detection in
poultry carcasses
IMAGE SENSORS:
 CMOS–lower-end applications
(e.g. mobile devices and small
cameras)
 CCD–cost-effective (in the
long-run) for use in Spectral
Imaging (e.g. Hyperspectral
Imagers, Spectrometers, X-ray
Imaging Systems)
Fig 8: Absorbance spectra of
whole corn kernel and aflatoxin-
contaminated kernel
(Source: Yao et al, 2015)
Fig 7: Reflectance spectra of
different types of vegetation
(Smith, 2012)

 http://www.nrcan.gc.ca/earth-sciences/geomatics/satellite-imagery-air-photos/satellite-imagery-products/educational-
resources/9393#answer1.
 https://en.wikipedia.org/wiki/Spectrograph.
 https://en.wikipedia.org/wiki/Charge-coupled_device.
 http://bwtek.com/spectrometer-introduction/.
 http://www.sensorsmag.com/machine-vision/growing-world-image-sensors-market-9533.
 https://en.wikipedia.org/wiki/Hyperspectral_imaging.
 Delwiche, S. (2015). Basics of Spectroscopic Analysis. In B. Park & R. Lu (Eds.), Hyperspectral Imaging Technology in
Food and Agriculture (pp. 57-79): Springer New York.
 Shaw, G.A., Burke, H.K. (2003). Spectral Imaging for Remote Sensing. MIT Lincoln Laboratory Journal. Available
online at: https://www.ll.mit.edu/publications/journal/pdf/vol14_no1/14_1remotesensing.pdf.
 Smith, R.B., 2012. Introduction to hyperspectral Imaging. Available online at:
http://www.microimages.com/documentation/Tutorials/hyprspec.pdf.
 Tian, H., 2000. Noise Analysis in CMOS Image Sensors. (Unpublished Thesis)
 Yao, H., Hruska, Z., Brown, R., Bhatnagar, D., Cleveland, T. (2015). Safety Inspection of Plant Products. In B. Park &
R. Lu (Eds.), Hyperspectral Imaging Technology in Food and Agriculture (pp. 127-172): Springer New York.

Multi spectral imaging sensors

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Multi spectral imaging sensors (20)

Recently uploaded (20)

Multi spectral imaging sensors

Editor's Notes