-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathM43_Multi.html
168 lines (130 loc) · 15.6 KB
/
M43_Multi.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
<html>
<head>
<p class="head">Spectral imaging</p>
<link rel="stylesheet" type="text/css" href="exampleStyle.css"/>
<hr>
</head>
<body>
<div id="Overview">
<p class="thema">Overview</p>
<p>Spectral imaging is a technique that combines <b>spectroscopy</b> and <b>imaging</b> to create a spectral map of a sample, which can provide information about the chemical and physical properties of the sample. It is a type of imaging that captures data from many different <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/15_Wave.html">wavelengths </a> of <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/1_Light.html">light</a> and gives a detailed view of an object or site.</p>
<p>Spectral imaging can be considered as a special case of spectroscopy ("<a class="def">imaging spectroscopy</a>") or a special case of imaging (“<a class="def">spectral imaging</a>”). Current practice is to use adjectives such as “<a class="def">multi-</a>” and “<a class="def">hyper-</a>” added to “spectral imaging” in order to characterize the number of used light wavelengths, that can vary from 10 to 2000 in one investigation.</p>
<div class="gallery">
<img src="Figures/hyperspec.jpg" width="400" height="100">
<div class="desc">Comparison of the image stacks in multispectral imaging, in which there are images taken in several different spectra, and hyperspectral imaging, in which there are images taken in many different spectra. Source: <a target="_blank" href="https://www.edmundoptics.de/knowledge-center/application-notes/imaging/hyperspectral-and-multispectral-imaging/">www.edmundoptics.de</a></div>
</div>
</div>
<div id="Tasks">
<p class="thema">Research tasks & applications</p>
<p>Spectral imaging is currently much used technique for a comprehensive examination of an object or material, providing insights into its composition, structure and condition, in different fields, such as chemistry, geology, medicine, food, agriculture, remote sensing and materials science. It is also an effective tool in conservation science allowing recording, analysis and protection of artifacts. It is becoming increasingly popular for use in cultural heritage sites around the world. </p>
<p>The tasks of multispectral imaging in applications for cultural heritage include: </p>
<ul class="b">
<li>to recover erased text in palimpsests; </li>
<li>to restore damaged manuscripts;</li>
<li>to characterize the paper components and their degradation or damage;</li>
<li>to document and analyze the various layers of a cultural site, such as paint layers on a wall or the layers of a burial mound;</li>
<li>to identify materials;</li>
<li>to detect anomalies. </li>
</ul>
</div>
<div id="P">
<p class="thema">Principle</p>
<p>The principle of multispectral imaging involves acquiring images of a scene or object at multiple <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/15_Wave.html">wavelengths</a> across the <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/13_EM_spectrum.html">electromagnetic spectrum</a>. This is done by taking multiple images of the same object from different wavelengths of <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/1_Light.html">light</a>. It allows for the analysis of different spectral bands to gain a comprehensive understanding of the material properties, composition, and characteristics. </p>
<p>Hyperspectral camera measures thousands or hundreds of thousands of spectra, instead of single <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/15_Wave.html">spectrum</a>. It is capturing images from a broad range of the electromagnetic spectrum – it can start with <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/11_ultraviolet.html">UV</a> light, extend through the <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/18_visible_light.html">visible</a> spectrum, and end in the near or short-wave <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/10_infrared%20light.html">infrared</a>. This extended wavelength range can reveal properties of material composition that are not otherwise apparent. The collected spectra are used to form an image of the target in a way that each image pixel includes a complete spectrum. By assigning positional data to the collected light spectrums, hyperspectral imaging provides 3D spectral data that is called <a class="def">data cube</a> or <a class="def">hypercube</a>, that is essentially a stack of images of the same object/scene taken on each wavelength. In a truly hyperspectral image, each pixel corresponds to coordinates, signal intensity, and wavelength information. Finally, there is an imaging with very dense and continuous spectral information for each pixel of the image. This allows to answer the questions “what”, based on the spectrum and “where”, based on the location. To manipulate hypercube, analysis <b>software</b> is required. </p>
<p><a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/31_Transmission.html">Transmissive</a> illumination images, <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/23_Reflection.html">reflection</a> images and <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/9_Luminescence_Fluorescence.html">fluorescence</a> images can be acquired in multi- and hyperspectral process. Usually, the combination of these three corresponding techniques (<a target="_blank" class="ref" href="https://cvertan.github.io/physics4dh.github.io/M43_3_Trans_Spec.html">transmittance spectroscopy</a>, <a target="_blank" class="ref" href="https://cvertan.github.io/physics4dh.github.io/M43_1_Reflectography.html">reflectography</a> and <a target="_blank" class="ref" href="https://cvertan.github.io/physics4dh.github.io/M43_2_Flu_Spec.html">fluorescence spectroscopy</a>) are used for research purposes. </p>
<p>The resulting multispectral data can be interpreted to gain insights into various aspects, such as material composition, <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/43_Ink_topology.html">ink topology</a>, surface characteristics, and hidden information. This analysis can involve comparing spectral signatures, applying machine learning algorithms, or referencing databases of known spectra for identification or classification purposes. </p>
<div class="gallery">
<img src="Figures/hyperspecIm.jpg" width="600" height="200">
<div class="desc"> Hyperspectral data cube with spectral signature. Source: <a target="_blank" href="https://www.youtube.com/watch?v=Zaw16EdM93g">Headwall Photonics</a>.
</div>
</div>
<div class="gallery">
<img src="Figures/multiS.jpg" width="400" height="100">
<div class="desc">Each pixel corresponds to coordinates, signal intensity, and wavelength information. Source: <a target="_blank" href="https://www.youtube.com/watch?v=ayp7hP0Xr8Q&t=142s">SpecimSpectral</a></div>
</div>
</div>
<div id="Equip">
<p class="thema">Used Equipment</p>
<p>The equipment used in spectral imaging can vary depending on the specific application, but the key components are the light source, spectrometer, camera and data analysis software. The equipment used in spectral imaging typically includes: </p>
<ol class="c">
<li><b>Light source</b> is used to illuminate the sample. Depending on the type of sample and the application, different types of light sources may be used, such as white light, filtered light or <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/22_Laser.html">laser light </a>. </li>
<li><b>Spectrometer</b> is used to separate the light from the sample into its component wavelengths (or frequencies). The most commonly used types of <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/27_Spectrometer.html">spectrometers</a> for spectral imaging are dispersive spectrometers and Fourier transform spectrometers.</li>
<li>Special <b>hyperspectral camera</b> (or microscope) is used to capture the spectral image of the sample. It typically includes a sensor or detector, each of which records the intensity of the <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/1_Light.html">light</a> at different <a target="frameterms" class="ref" href="https://cvertan.github.io/physics4dh.github.io/15_Wave.html">wavelength</a>.
</li>
<li><b>Data analysis software</b> can be used to create spectral maps, which show the distribution of different chemical or physical properties across the sample. Spectral imaging generates large amounts of data, and specialized software is required to analyze and interpret the data. </li>
</ol>
<div class="gallery">
<img src="Figures/Multi_S.jpg" width="400" height="100">
<div class="desc">Configuration of the hyperspectral imaging system. <a target="_blank" href=" "> </a></div>
</div>
<table>
<tr>
<th>Imaging system</th>
<!-- <th>Operating Range</th> -->
<th>Specification</th>
</tr>
<tr>
<td>EV™ multispectral imaging system </td>
<td><a target="_blank" href="https://mega-vision.com/products/ev-spectral-imaging-system/">www.mega-vision.com</a> </td>
</tr>
<tr>
<td>NASA’s Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) </td>
<td><a target="_blank" href="https://www.jpl.nasa.gov/missions/airborne-visible-infrared-imaging-spectrometer-aviris">www.jpl.nasa.gov</a></td>
</tr>
<tr>
<td>Digital microscope Dino-Lite </td>
<td><a target="_blank" href="https://www.dino-lite.com/">www.dino-lite.com</a></td>
</tr>
</table>
</div>
<div id="Cases">
<p class="thema">Case Studies</p>
<table>
<tr>
<th>Object studied</th>
<th>Tasks and procedure</th>
<th>Source</th>
</tr>
<tr>
<td>The Ethiopic overwritten manuscript Petermann II Nachtrag 24 from the Berlin State Library</td>
<td><p>This study will focus on the response of iron gall ink to multispectral illumination. The goal was to recover the erased text and make it legible for a scholar to read. </p>
<p>The multispectral imaging system developed by <a class="ref" target="_blank" href="https://mega-vision.com/products/ev-spectral-imaging-system/">MegaVision</a> (‘Archival and Cultural Heritage Imaging’) was used. Several LEDs of different wavelengths have turned on at the same time to generate white light for positioning the manuscript. During image capture, LEDs are turned on individually to illuminate the manuscript with only one wavelength of light, at a time. The camera has a 50 mega-pixel, panchromatic sensor that records the response of the manuscript leaf to each wavelength of light. Three types of optical images were captured: 1) simple reflectance of the light from the parchment surface; 2) fluorescence images under ultraviolet light illumination; 3) images taken of light that is transmitted through the parchment leaf. The leaf is illuminated from a light sheet that sits below the leaf. </p>
</td>
<td>[<a class="ref" href=bibliography.html#Knox_2020>Knox_2020</a>]</td>
</tr>
<tr>
<td>San Lorenzo Palimpsest (entitled <i>Campione dei Beni, 1504</i>), Archivio del Capitolo di San Lorenzo, Florence, Italy </td>
<td><p>The goal was to identify the compositions, recover the lost writings in manuscripts and create a publishable set of images that enhance the original notation. The task is made more difficult not only by the scraped content, but also because of the fact that by the end of the 15th century the manuscript had been completely disassembled, therefore the compositions appear in the wrong order today.</p>
<p>Manuscripts do not only include new readings for compositions known from other contemporary manuscripts, but also contains completely unknown compositions by Italian composers from the beginning of the 15th century. </p>
<p>The <a class="ref" target="_blank" href="https://mega-vision.com/products/ev-spectral-imaging-system/">EV™ multi-spectral imaging system</a> was used for imaging in this study employs a 50-megapixel monochromatic camera, 13 different wavelengths between 365 and 1050 nm, five filters and two raking lights in blue and infrared. The system is portable, so it was possible to transport it to the Archivio del Capitolo di San Lorenzo to image on site in one of the archive’s rooms prepared to accommodate all the equipment. This included a cradle specifically designed to hold delicate manuscripts. Since it was not possible to remove folios from the binding and produce images of each page individually, an acrylic plate was placed over each folio to keep it standing upright and to keep the book from closing. </p>
<p>Under illumination of the palimpsest with different light wavelengths, pictures of three types were taken by means of the camera using the optic filters:
1) raking-light images;
2) reflectance images;
3) fluorescence images.
</p>
<p>The use of raking lights allows the capture of topographical and texture information relating to the folio. Reflectance and fluorescence images allow to ‘show through’ over- and underwriting for the reader. </p>
<p>A total of 24 different reflectance and fluorescence images were taken and included in the processing work using the statistical methods of principal component analysis (PCA) and occasionally independent component analysis (ICA) implemented in the Excelis ENVI software package. </p>
</td>
<td>[<a class="ref" href=bibliography.html#Janke_2014>Janke_2014</a>]</td>
</tr>
<tr>
<td>Writing Materials in Geniza Fragments, Taylor-Schechter Collection at Cambridge University Library. </td>
<td>The goal was to identify ink typology. Researchers used a portable microscope (<a class="ref" target="_blank" href="https://www.dino-lite.com/">Dino-Lite</a>) with illumination in ultraviolet (390 nm), visible, and near infrared (940 nm) regions of the electromagnetic spectrum and magnifications of x50 to x200; micrographs to study the ductus of the script. </td>
<td>[<a class="ref" href=bibliography.html#Janke_2014>Cohen_2017</a>]</td>
</tr>
</table>
</div>
<p class="acknow">Acknowledgements:
[<a class="ref" href=bibliography.html#Janke_2014>Janke_2014</a>],
[<a class="ref" href=bibliography.html#Knox_2020>Knox_2020</a>],
[<a class="ref" href=bibliography.html#Marshall_2012>Polder_2021</a>],
[<a class="ref" href=bibliography.html#wiki>wiki</a>],
[<a class="ref" href=bibliography.html#Porebski_2022>Porebski_2022</a>],
[<a class="ref" href=bibliography.html#Yao_2022>Yao_2022</a>],
[<a class="ref" href=bibliography.html#Peery_2020>Peery_2020</a>],
[<a class="ref" href=bibliography.html#Cohen_2017>Cohen_2017</a>],
[<a class="ref" href=bibliography.html#edmundoptics>edmundoptics</a>],
[<a class="ref" href=bibliography.html#Gordon_2014>Gordon_2014</a>].</p>
<p class="important">More on subject: <a href="https://www.youtube.com/watch?v=_sUZ96YZOQU">www.youtube.com</a>, <a href="https://mixam.com/support/colour">www.youtube.com</a>, <a href="https://www.youtube.com/watch?v=GhpBmL5_OXw">www.youtube.com</a>.</p>
</body>
</html>