Hyperspectral Imaging (HSI) is a non-invasive technique that captures both spatial and spectral information across a wide range of wavelengths. It generates a high-dimensional dataset—called a spectral hypercube—where each pixel contains a full spectrum, enabling material identification without sampling or damaging the object. HSI is especially valuable in cultural heritage, as it reveals pigments, degradation products, and features invisible to the naked eye. By analyzing reflectance spectra, materials can be identified and changes such as discoloration or degradation tracked. This thesis presents the design and realization of a VIS–NIR–SWIR hyperspectral camera with extended spectral coverage. At its core is a dual-layer sensor: an InGaAs layer (1000–1700 nm) and a silicon layer (400–1000 nm). Combined, they span the 400–1700 nm range, covering the visible, near-infrared, and short-wave infrared regions. The sensor is coupled with an ultra-stable common-path birefringent interferometer (TWINS). Spectral information is retrieved by Fourier Transform of the interference pattern between temporally delayed replicas. The second chapter includes the characterization of sensor noise and the influence of lens-sensor coupling on field of view and working distance. The interferometer design is also analyzed, with discussion on vignetting and signal loss. The chapter concludes with an evaluation of the illumination strategy, crucial to cover such a broad spectral range. The third chapter presents hyperspectral measurements carried out on polymer samples to assess spectral discrimination, confirming the camera’s ability to distinguish different materials. Further tests were conducted on paintings, where areas that appeared identical in the visible range revealed differences in the infrared, highlighting material variations and underlying modifications made during the artistic process. These results demonstrate the system’s strong potential for broad-spectrum hyperspectral analysis, especially in the infrared range where many materials exhibit distinct signatures.
L’Hyperspectral Imaging (HSI) è una tecnica non invasiva che acquisisce simultaneamente informazioni spaziali e spettrali su un ampio intervallo di lunghezze d’onda. Genera un ipercubo spettrale, una struttura dati tridimensionale in cui ogni pixel contiene uno spettro completo, permettendo l’identificazione dei materiali senza campionamento o danneggiamento dell’oggetto. L’HSI è particolarmente utile nel campo dei beni culturali, poiché consente di rilevare pigmenti, prodotti di degrado e dettagli strutturali non visibili a occhio nudo. L’analisi degli spettri di riflettanza permette di identificare i materiali e monitorare alterazioni come scolorimenti o degrado chimico. Questa tesi presenta la progettazione e realizzazione di una camera iperspettrale VIS–NIR– SWIR con un’estesa copertura spettrale. Il sensore innovativo a doppio strato è costituito da uno in InGaAs (1000–1700 nm) e uno in silicio (400–1000 nm). La loro combinazione consente di coprire l’intervallo 400–1700 nm, includendo le regioni visibile, vicino e corto infrarosso. Il sensore è accoppiato a un interferometro birifrangente a cammino comune ultrastabile (TWINS). Le informazioni spettrali si ottengono tramite Trasformata di Fourier del pattern di interferenza tra due repliche temporali del segnale. Il secondo capitolo descrive la caratterizzazione del rumore del sensore e l’effetto del sistema ottico sulla distanza di lavoro e sul campo visivo. Si analizza anche la progettazione dell’interferometro, affrontando problematiche come vignettatura e perdita di segnale, concludendo con la scelta della strategia di illuminazione più adatta per l’intero intervallo spettrale. Il terzo capitolo riporta le misure iperspettrali su campioni polimerici, che dimostrano la capacità di distinguere materiali con firme spettrali differenti. Successivamente, sono stati analizzati alcuni dipinti: aree visivamente identiche hanno mostrato differenze nell’infrarosso, rivelando variazioni di materiali e modifiche apportate in fase esecutiva. I risultati dimostrano l’efficacia della camera per analisi iperspettrali su ampi intervalli, in particolare nell’infrarosso, dove molti materiali mostrano firme caratteristiche.
Design and development of a broadband hyperspectral imaging camera covering the VIS-NIR-SWIR spectral range
Mallia, Matteo
2024/2025
Abstract
Hyperspectral Imaging (HSI) is a non-invasive technique that captures both spatial and spectral information across a wide range of wavelengths. It generates a high-dimensional dataset—called a spectral hypercube—where each pixel contains a full spectrum, enabling material identification without sampling or damaging the object. HSI is especially valuable in cultural heritage, as it reveals pigments, degradation products, and features invisible to the naked eye. By analyzing reflectance spectra, materials can be identified and changes such as discoloration or degradation tracked. This thesis presents the design and realization of a VIS–NIR–SWIR hyperspectral camera with extended spectral coverage. At its core is a dual-layer sensor: an InGaAs layer (1000–1700 nm) and a silicon layer (400–1000 nm). Combined, they span the 400–1700 nm range, covering the visible, near-infrared, and short-wave infrared regions. The sensor is coupled with an ultra-stable common-path birefringent interferometer (TWINS). Spectral information is retrieved by Fourier Transform of the interference pattern between temporally delayed replicas. The second chapter includes the characterization of sensor noise and the influence of lens-sensor coupling on field of view and working distance. The interferometer design is also analyzed, with discussion on vignetting and signal loss. The chapter concludes with an evaluation of the illumination strategy, crucial to cover such a broad spectral range. The third chapter presents hyperspectral measurements carried out on polymer samples to assess spectral discrimination, confirming the camera’s ability to distinguish different materials. Further tests were conducted on paintings, where areas that appeared identical in the visible range revealed differences in the infrared, highlighting material variations and underlying modifications made during the artistic process. These results demonstrate the system’s strong potential for broad-spectrum hyperspectral analysis, especially in the infrared range where many materials exhibit distinct signatures.| File | Dimensione | Formato | |
|---|---|---|---|
|
2025_Mallia_Tesi_01.pdf
solo utenti autorizzati a partire dal 02/07/2026
Descrizione: Testo della Tesi
Dimensione
18.69 MB
Formato
Adobe PDF
|
18.69 MB | Adobe PDF | Visualizza/Apri |
|
2025_Mallia_Executive Summary_02.pdf
solo utenti autorizzati a partire dal 02/07/2026
Descrizione: Executive Summary
Dimensione
2.22 MB
Formato
Adobe PDF
|
2.22 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/239910