ECEU 692 Subsurface Imaging Course Notes Part 3

ECEU 692 Subsurface Imaging Course Notes Part 3: Imaging with Light (2) Hyperspectral Imaging Profs. Brooks and Di. Marzio Northeastern University Spring 2004 January 2004 Chuck Di. Marzio, Northeastern University 1

Hyperspectral Imaging Concepts ER-2 • • Absorption Spectroscopy Scattering Reflectance Spectroscopy • Imaging • Hyperspectral Imaging • The Data Cube January 2004 Chuck Di. Marzio, Northeastern University AVIRIS 2

Beer’s Law January 2004 Chuck Di. Marzio, Northeastern University 3

ka. Specific Absorption, /cm /M Sample Absorption Spectra Data from Van Assendelft, O. W. , Spectrophotometry of Haemoglobin Derivatives, Springfield, IL: Thomas, 1970. Hb 686 January 2004 Hb. O 2 l, Wavelength, nm Chuck Di. Marzio, Northeastern University Graph by Di. Marzio, Warner, Dwyer, Northeastern University, 1994 4

Light in a Turbid Medium • Specular Components – Fresnel Coefficients – Beer’s Law • Diffuse Components – More Complicated • Some Examples – Paint – Clouds January 2004 Chuck Di. Marzio, Northeastern University 5

Reflectance Spectroscopy; Weak Scattering January 2004 Chuck Di. Marzio, Northeastern University 6

A Simple Model ma(l 1)=C 1 k 1(l 1)+ C 2 k 2(l 1) ma(l 2)=C 1 k 1(l 2)+ C 2 k 2(l 2) Two Measurements, Two Equations, and Two Unknowns Later, Prof. Brooks will talk about inverting these equations January 2004 Chuck Di. Marzio, Northeastern University 7

Reflectance Spectroscopy; Strong Scattering No Hb Hb. O 2 Hb January 2004 Chuck Di. Marzio, Northeastern University 8

Pulse Oximeter • 2 LED’s – Different IR Wavelengths • DC Components – Total Hb, Hb. O 2, Other Absorbers • Pulsatile Component – Arterial Hb, Hb. O 2 January 2004 Chuck Di. Marzio, Northeastern University 9

Hyperspectral Imaging Overview • • Definition (Multi-, Hyper, Ultra-Spectral) The Data Cube Image Acquisition Hardware, Data Visualization HOSIS - Hyperspectral Oxygen Saturation Imaging System: – CCD – Lyot Filter – Digital Imaging January 2004 Chuck Di. Marzio, Northeastern University 10

Hyperspectral Imaging • Definitions? y l x – Imaging in multiple bands with width less than (10%? ) of the center frequency – Imaging in more than (25? ) bands – Imaging in Contiguous Bands The Hyperspectral “Data Cube” Chuck Di. Marzio, Northeastern University January 2004 11

Hyperspectral Imaging Hardware Concepts (1) • Camera Coordinates are Spatial • Successive Images are Different Wavelengths y l x Tunable Filter January 2004 B/W Camera Chuck Di. Marzio, Northeastern University 12

Hyperspectral Imaging Hardware Concepts (2) • Pushbroom Scan – Airborne or Vehicle Mounted? – Camera Coordinates are y, l – Successive Scans are Different values of x y l x January 2004 Chuck Di. Marzio, Northeastern University 13

HOSIS Instrumentation Block Diagram Processed Images Halogen Lamp CCD Camera Tunable Filter See Supplementary Notes Computer System with Video Board Target Tissue January 2004 Chuck Di. Marzio, Northeastern University Thanks to Peter Dwyer at Lucid Technologies, MGH, and NU 14

Quantitative Imaging • Quantitative Calculations • Sources of Variation – Difficult – Subject to Change • Calibration Standards – Light Level – Reflectance January 2004 – – – Light Source Camera Sensitivity Filter Losses Geometry Atmosphere? Other? Chuck Di. Marzio, Northeastern University 15