VideometerLab 3 Multi-Spectral Imaging

Transcription

1 analytikltd VideometerLab 3 Multi-Spectral Imaging Rapid Non-destructive Analysis of Heritage Artefacts Adrian Waltho, Analytik Ltd (Cambridge, UK) adrian.waltho@analytik.co.uk

2 Normal Colour Imaging Traditional colour imaging uses three broad bands of colour: Red, Green and Blue

3 Normal Colour Imaging Chloro-a High Low High Chloro-b High Low High RGB photographs have limited spectral resolution Chlorophyll a and b give almost the same RGB signal and are not spectrally separated

4 Multispectral Imaging Chloro-a High Low Low High Chloro-b Med High Med Low Using just 4 wavelength bands with tightly defined ranges, Chlorophyll a and b can easily be distinguished VideometerLab 3 uses 19 wavelength bands

5 Multispectral Imaging Ultraviolet Near-Infrared Infra-Red Red Yellow Green Blue Ultra-Violet Many images obtained at selective wavelength bands Each image pixel contains spectral data points Spectral signature reveals chemo-specific information See spatial location of surface chemical variation

6 VideometerLab 3 Schematic Narrowband illumination provided by 19 LEDs between 375nm-970nm (UV-Vis-NIR) Integrating sphere diffuses light onto sample 2056x2056 CCD camera captures reflectance at each LED wavelength for each pixel Emission filter wheel for longpass/bandpass filtering Bright-field or dark-field lighting for translucency or diffuse scatter imaging Camera Integrating sphere LEDs Sample Emission filter wheel

7 VideometerLab 3 Schematic Precise lighting control for optimal illumination of each sample across whole spectrum Diffuse light on sample ensures even lighting from all directions Broadband camera spectrally and spatially calibrated (NIST) Emission filters enable multispectral reflectance and fluorescence imaging abilities Powerful, user-friendly image analysis and quantification software for data and modelling Camera Integrating sphere LEDs Sample Emission filter wheel

8 VideometerLab 3 Schematic

9 VL3 Multispectral Imaging of Heritage Items

10 VL3 Multispectral Imaging of Heritage Items It is critical that any analysis techniques used on historical artefacts are entirely nondestructive Multispectral imaging can be used safely to analyse and investigate historical artefacts Fast, non-destructive and objective method to reveal new information Complete record of illumination settings and image data allows absolute tracking of colour changes, corrosion etc. over time A number of historical texts and objects have been investigated using the VideometerLab 3

11 VL3 Imaging of Elder Westrogothic Law National Library of Sweden

12 VL3 Imaging of Elder Westrogothic Law National Library of Sweden

13 VL3 Imaging of Elder Westrogothic Law National Library of Sweden

14 VL3 Imaging of Elder Westrogothic Law National Library of Sweden

15 VL3 Imaging of Elder Westrogothic Law National Library of Sweden

16 VL3 Imaging of Elder Westrogothic Law National Library of Sweden During the work with the Elder Westrogothic Law (ÄldreVästgötalagen) we used the VideometerLab 3 to make spectral imaging and analysis of four partially or largely unreadable pages. During this process we obtained valuable new information and the VideometerLab 3 proved to be a highly useful and efficient tool. Thanks to the Videometer images, we now know a lot more about the author than we knew before. -- Per-Axel Wiktorsson, Professor emeritus in Swedish language, Örebro University

17 Bodleian Library Papyrus Hieroglyphics Full area of VL3 image of papyrus srgb photographic image based on spectral response at red, green and blue wavelengths only

18 Bodleian Library Papyrus Hieroglyphics Detail of VL3 image of papyrus srgb photographic image based on spectral response at red, green and blue wavelengths only

19 Bodleian Library Papyrus Hieroglyphics Same area at 850nm infrared illumination only False coloured to highlight contrast to our eye Blue coloured pixels have low 850nm reflectance, red coloured pixels have high 850nm reflectance Note smudge (circled), invisible in srgb image fingerprint contamination?

20 Bodleian Library Papyrus Hieroglyphics Same area after transformation with statistical processing ncda algorithm is trained to look for smudge spectrum pixels Lighter pixels are more similar to the smudge spectrum, darker pixels are more dis-similar No other similar smudges are apparent in this field of view

21 Bodleian Library Palimpsest Analysis Palimpsest, full field of view, srgb image Original text has been overwritten with darker ink, making the original text hard to read

22 Bodleian Library Palimpsest Analysis Palimpsest, zoomed field of view, srgb image Original text has been overwritten with darker ink, making the original text hard to read

23 Bodleian Library Palimpsest Analysis Same area after transformation with statistical processing ncda algorithm is trained to look for original ink pixels Lighter pixels are more similar to the ink spectrum, darker pixels are more dis-similar Original ink writing is highlighted to make reading easy

24 Historic Scotland Pigment analysis Circled are cadmium red and lead white pigment samples

25 Historic Scotland Pigment analysis Circled is a mix of cadmium red and lead white

26 Historic Scotland Pigment analysis The VideometerLab can build up a library of the spectral responses of each pigment

27 Historic Scotland Pigment analysis The VideometerLab can then recognise this spectrum in unknown samples. We see that the pure cadmium red and the cadmium red mix are both hits for the cadmium spectrum

28 Historic Scotland Pigment analysis We can also see that the more pure spectrum registers as a closer fit than the cadmium-lead mix. We see this because the pure cadmium is much more orangey in the false-colour image

29 Historic Scotland Pigment analysis This technique can be applied to paintings, for example this portrait of a captain from Leith

30 Historic Scotland Pigment analysis Using false-coloration, we can see that the yellow pigment used on the captain s hatband is the same (or very similar) to the yellow on his buttons, but different from the yellow on his braiding

31 Royal Armouries Fencing Manual

32 Royal Armouries Fencing Manual

33 Royal Armouries Fencing Manual This image shows the redness on the left fighter s cheek is spectrally the same as the scribbling in the middle of the picture and the spot in the bottom left

34 Royal Armouries Fencing Manual The VL3 can prove this quantitatively as well the next slide shows the spectral properties of the points in the center of the small crosses below

35 Royal Armouries Fencing Manual You can see that 3 and 4 (teal and brown, page and sword) are significantly different from 1 and 2 in terms of their spectral profile.

36 Royal Armouries Sten Gun Visually, the Sten gun looks scratched randomly. In fact, it is known there is an engraving on the side

37 Royal Armouries Sten Gun The VL3 can isolate and enlarge the engraving

38 Royal Armouries Sten Gun The VL3 can also spectrally enhance the contrast

39 Royal Armouries Sten Gun Ordinary RGB imaging cannot do this - it is only possible to get the clarity of the enhanced image at 850 nanometers where the scratches start to respond differently to the metal under illumination, and RGB cameras only go up to around 700nm

40 Royal Armouries Sten Gun The VideometerLab software can false-colour images to aid understanding

41 Royal Armouries Sten Gun It can also focus tightly on specific sections of the image to contrast them better it is possible to improve signal to noise almost without limit

42 Royal Armouries Sten Gun ΧΡΙΣΤΗΦΗΣ ΗΛΙΑ Christofis Elias

43 Royal Armouries Sten Gun ΛΟΧΟΣ ΣΤ Sixth Company

44 Royal Armouries Sten Gun ΛΕΥΚΩΣΙΑ Lefkosia

45 Royal Armouries Sten Gun The gun belonged to a Greek Cypriot named Elias Christofis who fought in World War II. (Lefkosia is the capital of Cyprus) Some of the decisions between similar looking letters might appear odd, but make a lot of sense when looked at multispectrally: For example, this letter looks like it could easily be a lambda or an aleph, but when viewed at 920 rather than 850nm, the line in the middle disappears

46 Museum Conservation Service Faded Elizabeth R signature on official declaration of war communication document to overseas UK forces (srgb image)

47 Museum Conservation Service Yes Faded Elizabeth R signature on official declaration of war communication document to overseas UK forces (srgb image) Highlight training set of pixels to teach the software the spectral pattern of interest No

48 Museum Conservation Service Faded Elizabeth R signature on official declaration of war communication document to overseas UK forces Highlight training set of pixels to teach the software the spectral pattern of interest Software scores every pixel in the image using the trained ncda algorithm ncda score determines pixel false colour value Signature s spectrum is highlighted against all other dissimilar spectra on the page

49 Museum Conservation Service Faded Officer s co-signature on official declaration of war communication document to overseas UK forces (srgb image)

50 Museum Conservation Service No Faded Officer s co-signature on official declaration of war communication document to overseas UK forces (srgb image) Highlight training set of pixels to teach the software the spectral pattern of interest Yes

51 Museum Conservation Service Faded Officer s co-signature on official declaration of war communication document to overseas UK forces Highlight training set of pixels to teach the software the spectral pattern of interest Software scores every pixel in the image using the trained ncda algorithm ncda score determines pixel false colour value Signature s spectrum is highlighted against all other dissimilar spectra on the page

52 Museum Conservation Service 1858 written in pencil Faded pencil inscription on the rear of a photograph of unknown British officer in 1858 (srgb image, date circled) Highlight training set of pixels to teach the software the spectral pattern of interest

53 Museum Conservation Service Faded pencil inscription on the rear of a photograph of unknown British officer in 1858 Highlight training set of pixels to teach the software the spectral pattern of interest Software scores every pixel in the image using the trained ncda algorithm ncda score determines pixel false colour value Pencil spectrum is highlighted against all other dis-similar spectra on the page

54 Museum Conservation Service Suspected signature painted over in final composition Almost invisible in srgb image

55 Museum Conservation Service Suspected signature painted over in final composition Almost invisible in srgb image Highlight training set of pixels to teach the software the spectral pattern of interest

56 Museum Conservation Service Suspected signature painted over in final composition Almost invisible in srgb image Highlight training set of pixels to teach the software the spectral pattern of interest Software scores every pixel in the image using the trained ncda algorithm ncda score determines pixel false colour value Signature s spectrum is highlighted against all other dissimilar spectra on the page

57 Multispectral Image Analysis A variety of image processing tools are available in VideometerLab s analysis software Analysis protocols and models can be saved and applied to new samples for routine analysis Multiple analysis models can be run on the same image data to quantify separate parameters Analysis can be batch-run to aid in high-throughput data acquisition Image archiving provides a record of samples for future data mining

58 VL3 Multispectral Imaging Nuanced, human-like analysis of historical artefacts Sophisticated data analysis tools for the average user Objective, repeatable data collection with easy export Rapid, non-destructive investigation and archiving Multispectral imaging at Analytik.co.uk VL3 Datasheet PDF