Upload
nguyennhu
View
213
Download
0
Embed Size (px)
Citation preview
ECE 5616 OE System Design
Robert McLeod 18
Geometrical opticsDesign of imaging systems
“Geometrical optics is either very simple, or else it is very complicated”
Richard P. Feynman
What’s it good for?1. Where is the image?2. How large is it?3. How bright is it?4. What is the image quality?1
“Useful when size of aperture is > 100 ”
Why?
1P. Mouroulis & J. Macdonald, Geometrical Optics and Optical Design, Oxford, 1997
•Design of ideal imaging systems with geometrical optics
ECE 5616 OE System Design
Robert McLeod 19
The ray equationApprox. solution of ME when n(r) is slow
rSjkerrE
0EAssume slowly varying
amplitude E and phase S
rnrS 22
Substitute into isotropic wave equation and retain lowest terms…
zkykxkrSk zyx
0 E.g. plane wave
22200 zyxkrSk
E.g. spherical wave
Countours of S(r) at multiples of 2
Sn(r)
S(r) Optical path length [m]
Ray equation – reduction of Maxwell’s equations.
“Ray” = curve to S(r)
•Design of ideal imaging systems with geometrical optics–Ray and eikonal equations
dsrnB
A
ECE 5616 OE System Design
Robert McLeod 20
The eikonal equationAn equation for evolution of ray trajectory
ds
rdrnsrnrS
ˆ
s Parametric distance along ray [m]Ray trajectory [m] sr
ds
rdrn
ds
drS
ds
d
then take derivative in s,
and chain rule for ,
rSrn
rSrS
ds
rdrS
ds
d
finally apply another identity.
rnrnrn
rSrSrn
2
2
1
2
1
rnds
rdrn
ds
d
•Design of ideal imaging systems with geometrical optics–Ray and eikonal equations
Take square-root of ray equation,
r
rdr
dss
sd
rd
ˆ
Ray
Saleh & Teich 1.3, Born & Wolf 7th edition, pg. 129
ECE 5616 OE System Design
Robert McLeod 21
Ray trajectory n=constantFinally!
rnds
rdrn
ds
d
•Design of ideal imaging systems with geometrical optics–Ray and eikonal equations
Eikonal
000
nds
rdn
ds
d
scsr
n(r) = no, a constant
where is a constant.c
Thus, we have discovered that, in homogenous materials, rays travel in straight lines!
ECE 5616 OE System Design
Robert McLeod
History of geometrical optics
22
•Design of ideal imaging systems with geometrical optics–Rays and refraction
• 280BC Euclid's Catoptrics light goes in straight line. However light was emitted by eye (not)
– law of reflection was correctly formulated in Euclid's book
• Hero of Alexandria, in his Catoptrics (first century BC), also maintained that light travels with infinite speed.
– adopted the rule that light rays always travel between two points by the shortest path
• 1000 AD Arab philosopher “Alhazan” (Abu'ali al-hasan ibn al-haytham) first person to realize that light actually travels from the object seen to the eye
• 1676 Olaf Römer demonstrated that light must have a finitevelocity, using his timings of the successive eclipses of the satellites of Jupiter
• 100-170AD Claudius Ptolemy rough law of refraction was studied experimentally (only close to normal incidence)
• 1621 Dutch mathematician Willebrord Snell comes up with correct law using sines. Does not publish
• 1637 French philosopher René Descartes was the first to publish, French countries call Snell’s law Descartes law of refraction
• 1658 French mathematician Pierre de Fermat demonstrated that all three of the laws of geometric optics can be accounted for on the assumption that light always travels between two points on the path which takes the least time
Kevin Curtis OESD lecture notes
ECE 5616 OE System Design
Robert McLeod 23
Pinhole camera (1/2)
•Design of ideal imaging systems with geometrical optics–Rays and refraction
• Mo Ti, China, invented 5th century BC• Aristotle observes image of eclipse cast through leaves, 300 BC
• Aristotle formulates theory of light and color (some right, some wrong)• Alhazen (Ibn Al-Haytham) invented CA 1000 AD• Della Porta invented CA 1600• Kepler named it Camera Obscura and suggested lens for efficiency, 1604
l
l’
• Magnification, M = - l’ / l• Very large depth of focus• Very small chromatic effects• Very poor power efficiency• Dr. Webster Cash at CU
funded by NASA Institute for Advanced Concepts to examine pinhole camera in space as extrasolar planet imager.
Abelardo Morell pinhole image of Manhattan
ECE 5616 OE System Design
Robert McLeod
Pinhole camera (2/2)Impact of pinhole size
24
•Design of ideal imaging systems with geometrical optics–Rays and refraction
• A pinhole camera has nolens but uses a very smallhole some distance fromthe film/screen to producean image. If we assumethat light travels in straightlines, then the image of adistant point source will bea blur whose diameter isthe same as the pinhole.However, diffraction willspread the beam into theAiry disk.
• Image is sharpest when the geometrical blur (the pinholesize) is equal to the diffraction blur (Airy disk diameter).
• Design example: What pinhole diameter is optimum for avisible-light camera that is 100 mm deep?
Geometrical blur
Diffraction blur
D/2
D/2
L=100 mm
mm 37.0
1001055.44.2
44.2
/2/22.1
sin22.1
3
LD
LDD
Kevin Curtis OESD lecture notes
ECE 5616 OE System Design
Robert McLeod
Evolution of the eye
25
•Design of ideal imaging systems with geometrical optics–Rays and refraction
Wikipedia
Lens evolved to improve radiometric efficiency of an otherwisepretty good imaging system.
ECE 5616 OE System Design
Robert McLeod 26
Fermat’s principleAnother (important) form of the eikonal
• Propagation time = S / c• Hero of Alexandria (in Catoptrica, ca 50
AD): “Light travels in straight lines”• Pierre de Fermat (ca 1650): “Light travels the
path which takes the minimum time.”• Correct: “The time of travel is stationary:”
– Example 1: Concave mirror with radius of curvature < ellipse
– Example 2: Lifeguard problem– See Born and Wolf for a derivation from
Maxwell’s equations
• Why do we care? At an image point, all OPL must be equal.
dsrnSB
A
lengthpath Optical
0 dsrnB
A
P. Mouroulis & J. Macdonald, Geometrical Optics and Optical Design, pg 11-12, Oxford, 1997R. Feynman, Lectures on Physics
•Design of ideal imaging systems with geometrical optics–Fermat’s principle
ECE 5616 OE System Design
Robert McLeod
The lifeguard problem
27
•Design of ideal imaging systems with geometrical optics–Fermat’s principle
A
BSand
n
Watern´
Problem: The lifeguard atpoint A needs to get to thedrowning swimmer at B inminimum time. What is theoptimum point at which toenter the water given that thelifeguard runs at speed V/n onsand and swims at speed V/n´in water?
Solution 1: Write equation fortravel time with entry point asparameter, minimize time,solve for entry point.
Example: If lifeguard onshore, this is a problem oftotal internal reflection, so ´is the critical angle:
´
A
BSand
n
Watern´
´
n
n
1-sin90°
Solution 2: Use Fermat’sprinciple which says thatangles obey Snell’s Law, usethis constraint to solve forpath.
ECE 5616 OE System Design
Robert McLeod 28
RadiometryReview of terminoligy & inverse square law
•Design of ideal imaging systems with geometrical optics–Radiometry
O’Shea chapter 3, Mouroulis and Macdonald 5.3
Q Energy [J] Power (flux) [J/s=W]I Intensity [W/sr]E Irradiance [W/m2]L Radiance (photometric “brightness”) [W/(sr m2)]
2R
A Solid angle is area of sphere
subtended A over radius of sphere R2
I Intensity I of point source is power emitted into solid angle
2R
IA
E
Irradiance E on surface is
power per unit area A
Irradiance of surface by a point sourceis given by intensity I of point source
over distance to surface R2
AL
Radiance of surface normal area A illuminated by a point source of power
radiating into a solid angle
ECE 5616 OE System Design
Robert McLeod
Examples
29
•Design of ideal imaging systems with geometrical optics–Radiometry
100 W
2 m 222 m
W25
m 2
W100
R
IE
What is irradiance 2 m from 100 W light bulb?
2 m
What is brightness of a typical 100 W light bulb?
Sr4mareafilament
W2
L W/m2/Sr] for a bulb
with a 1 mm2 filament
What is brightness of a typical 1 W laser?
Lots of power, moderate area, but huge radiation angle = low brightness.
2
22220
2220
1
radm
W
radangle divergencem
W
0
MMw
MwL
w
W/m2/Sr] for aperfect 1 W laser at =1 um
Less power and similar area to the lamp filament, but MUCH smaller radiation angle = high brightness. Besides the potentially narrow spectrum, brightness is what sets lasers apart from lamps.
ECE 5616 OE System Design
Robert McLeod 30
Radiometry via rays
•Design of ideal imaging systems with geometrical optics–Radiometry
Consider a finite source radiating into a cone
A1 A2
R2
R1
24 RAE
Now launch a set of rays inside this cone and examine the ray density as a function of radius
ER
N
A
N
24
A general proof that ray density is proportional to irradiance can be found in Born and Wolf.
Irradiance E on surface Agiven power into cone
Mouroulis and Macdonald 1.4
ECE 5616 OE System Design
Robert McLeod 31
Postulates of geometrical optics
• Rays are normal to equi-phase surfaces (wavefronts)
• The optical path length between any two wavefronts is equal
• The optical path length is stationary wrt the variables that specify it
• Rays satisfy Snell’s laws of refraction and reflection
• The irradiance at any point is proportional to the ray density at that point
•Design of ideal imaging systems with geometrical optics–Foundations