Measuring Light d dA (solid angle subtended by ) - PowerPoint PPT Presentation

12/6/16 Radiometric Terms Measuring Light ω d dA (solid angle subtended by ) source dA ' R ω d θ (foreshortened area) Lighting θ r i dA dA (surface area) Surface Irradiance Surface Radiance d 2 Φ = Φ L (watts / m 2 - d = θ ω E ( dA cos r ) d steradian ) ( watts / m 2 ) dA • Flux emitted per unit foreshortened area per unit solid angle θ Surface Camera • Light Flux (power) incident per unit surface • L depends on direction r area coming from a hemisphere of directions • Surface can radiate into whole hemisphere. • Does not depend on where the light is Image Quantization discretizes scene radiance at each • L depends on reflectance properties coming from of surface. pixel into a “brightness” or color value Digital Image Quantization Photometric Terms • A well-exposed photograph has a histogram that has values close to 0 near the minimum and • Photometry is the measurement of light as maximum brightness values so as not to lose detectable by the human eye information • Just like radiometry except weighted by the – No saturated (over-exposed) regions spectral response of the eye – No dark (under-exposed) regions • Luminance is the analog of radiance • Brightness values representing smoothly – Measured in lumens/m 2 -steradian (= nit) changing radiance should not be noticeable • Illuminance is the analog of irradiance – Too few gray levels leads to false contours in areas of – Measured in lumens/m 2 (= lux) the image where brightness changes slowly • Most digital cameras represent brightness by 8 bits per pixel, or 8 bits per color (R,G,B) 1

12/6/16 Dynamic Range False Contours Dynamic range is a measure of the contrast in the scene, i.e., the ratio between the brightest area and darkest area Real-World Scene Dynamic Range Real-World Scene Dynamic Range is Often Large is Often Large 1 : 1 • Some luminance levels of real scenes – Starlight 10 -3 cd/m 2 1,500 : 1 – Moonlight 10 -1 – Indoor lighting 10 2 – Sunlight 10 5 25,000 : 1 • Contrast ratio often 10,000 : 1 in a scene 400,000 : 1 2,000,000,000 : 1 2

12/6/16 Dynamic Range of Dynamic Range of Image Sensors Various Display Devices • Ratio of maximum possible signal (“full well capacity”) and total noise signal in the dark • LCD display 700 : 1 • Print film 128 : 1 • Common CCD sensors have dynamic range around 4,000 : 1, requiring 12 or 13 bits per pixel • Color negative 256 : 1 • A/D converters have conversion uncertainty that reduces • Positive slide 4,096 : 1 the usable dynamic range by ~1 bit • Human eye 100 : 1 static dynamic range • High-end CCDs have larger dynamic range • Human eye 1,000,000 : 1 dynamic range by • CMOS sensors have lower dynamic range adapting exposure geometrically and chemically to allow a sensitivity of ~10 -6 to 10 6 cd/m 2 • Bottom line: Image sensor dynamic range is not high enough to capture high dynamic range scenes Method 1: Limit Dynamic Range Image Dynamic Range is Too Small • W. Eugene Smith photo of Albert Schweitzer • Today’s cameras: limited dynamic range (LDR) • Overexpose bright areas • Correctly expose dark areas • 5 days to print High Exposure Image Low Exposure Image • We need 5-10 million values to store all brightnesses around us • But, typical 8-bit cameras provide only 256 values 3

12/6/16 Long Exposure Short Exposure 10 -6 10 6 10 -6 10 6 High dynamic range High dynamic range Real world Real world 10 -6 10 6 10 -6 10 6 Image Image 0 to 255 0 to 255 Method 2: Contrast Reduction How Humans Deal with Dynamic Range • Match limited contrast of the medium • We're sensitive to contrast (multiplicative) • Preserve details – A ratio of 2 : 1 is perceived as the same contrast as a ratio of 200 : 100 – Illumination has a multiplicative effect 10 -6 10 6 High dynamic range – Use the log domain as much as possible Real world • Dynamic adaptation (very local in retina) – Pupil – Neural – Chemical 10 -6 10 6 • Different sensitivity to different spatial Image frequencies Low contrast 4

12/6/16 Contrast Reduction: Fill-in Flash Perceived Brightness is Non-Linear • Use a flash to reduce contrast Exposure for outside Exposure for inside Average exposure Using fill flash From Le Livre de la Photo Couleur (Larousse) Filtering: Black and White Graduated Neutral Density Filtering Red/orange/yellow filters darken the sky • Art Wolfe: “In the late evening light, I composed this image using a graduated neutral-density filter to bring the overall exposure into alignment, thus preserving the detail in the clouds in the sky and the reflections on the water.” No filter With red filter Source: Ansel Adams http://www.artwolfe.com/ 5

12/6/16 Dodging and Burning Dodging and Burning • Must be done for every single print! • During print making process • Hide part of the print during exposure – Manually select areas to increase or decrease exposure Straight print After dodging and burning From The Master Printing Course, Rudman HDR Examples High Dynamic Range Digital Imaging Idea: Take multiple photos to cover the full dynamic range, then combine best parts from each photo John Adams 6

12/6/16 John Adams Alberto Carrozzo Relationship between Scene Radiance In-Camera Digital Pipeline and Image Brightness • Before light hits the image plane: • Photosites transform photons into charge (electrons) Scene Image Scene Lens – Sensor itself is linear Radiance, L Irradiance, E • Then goes through analog-to-digital (A/D) Assume known mapping converter • After light hits the image plane: – Usually 12 or 14 bits/channel • Stop here when shooting in RAW mode Camera Image Measured • Then image processing and a “response curve” Electronics Irradiance, E Pixel Values, Z are applied Non-linear mapping • Quantized and saved as 8-bit JPEG Can we go from measured pixel value, Z , to scene radiance, L ? 7

12/6/16 Relation between Pixel Value, Z , and Camera is Not a Photometer Image Irradiance, E Image Camera Measured • Limited dynamic range Electronics Irradiance E Pixel Values, Z – Use multiple exposures The Camera Response Function relates image irradiance , • Unknown, nonlinear response E , at the image plane to the measured pixel intensity value, Z → g : E Z – Difficult to convert pixel values directly to radiance • Solution: – Recover radiometric response curve from multiple exposures, then reconstruct the radiance map (Grossberg and Nayar) g is monotonic and smooth for all cameras High Dynamic Range (HDR) Imaging Ways to Vary Exposure 1. Capture multiple (usually 3 or 5) images with § Shutter speed different exposure settings § F-stop (aperture) 2. Estimate the Camera Response Function 3. Estimate radiance map : for each pixel, combine § Neutral Density (ND) filters the calibrated images (for example, by weighted § ISO / ASA averaging) Exposure X = E Δ t = 4. Tone map the image into a displayable range irradiance × time ⇒ Halving E and doubling Δ t will not change exposure – reciprocity property (Mitsunaga) 8

12/6/16 Exposure Value (EV) How to Best Vary Exposure? • Combinations of shutter speeds and apertures that give the same exposure (= irradiance * • Varying aperture changes the depth of field and time) (reciprocity property) can cause vignetting • Varying ISO changes the “graininess” • Each increment in • 1/1000 @ f/1.4 speed or aperture is a • Varying shutter speed is best • 1/500 @ f/2 “stop” • 1/250 @ f/2.8 • 1/125 @ f/4 • Assume scene is static, camera is static, and • Similarly, doubling the • 1/60 @ f/5.6 ISO value means it is lighting is static, so all images will be in register • 1/30 @ f/8 twice as sensitive, so 1/125 @ ISO 400 = • 1/15 @ f/11 1/250 @ ISO 800 • 1/8 @ f/16 • 1/4 @ f/22 Step 1: Capture Images with Different Shutter Speed Shutter Speeds (Varying Exposure) • Ranges: Canon D30: 30 to 1/4,000 sec Sony VX2000: ¼ to 1/10,000 sec • Pros: – Directly varies the exposure – Usually accurate and repeatable • Issues: – Noise in very long and very short exposures Assume scene is static, camera is static, and lighting is static, so all images are in register 9

12/6/16 Multiple Exposure Photography Capturing a Set of Images • Sequentially measure all segments of the range • Commonly, take 3 images at -2 EV, 0 EV, +2 10 -6 10 6 High dynamic range EV, or 5 images at increments of 1 EV Real world • Many modern digital cameras have “ AEB mode ” meaning “ automatic exposure 10 -6 10 6 bracketing ,” which automatically takes Image (usually) 3 photos at ±2 EV Low contrast • Expensive DSLR cameras can bracket 3, 5, 7, 9 images at various EV increments Multiple Exposure Photography Multiple Exposure Photography • Sequentially measure all segments of the range • Sequentially measure all segments of the range 10 -6 10 6 10 -6 10 6 High dynamic range High dynamic range Real world Real world 10 -6 10 6 10 -6 10 6 Image Image Low contrast Low contrast 10