Digital Image Processing: A Remote Sensing PerspectiveFW 5560Lecture 4Initial Display Alternatives and Scientific VisualizationInitial Display Alternatives and Scientific VisualizationDisplaying and analyzing remotely sensed data involves scientific visualization, which is definedas: “visually exploring data and information in such a way as to gain understanding and insight into the data”. The difference between scientific visualization and presentation graphics is that the latter are primarily concerned with the communication of information d lt th t l d d t d B i i tifi i li tiand results that are already understood. By using scientific visualization we want to understand the data and gain insight.All raster images that have a pixel brightness value at each row and column in a matrix arebitmappedimages The tonerow and column in a matrix are bitmappedimages. The tone or color of the pixel in the image is a function of the value of the bits or bytes associated with the pixel, and the i l ti th t t k l illktblmanipulation that takes place in a color look-up table.For example, the simplest bitmapped image is a binary image consisting of just ones (1) and zeros (0).Ch t i ti f ACharacteristics of A Binary Bitmapped ImageRGB Color Coordinate SystemDigital remote sensor data are usually displayed using a Red-Green-Blue (RGB) color coordinate system based on additive l th d th th i l f d d blcolor theory and the three primary colors of red, green, and blue.Additive color theoryis based on what happens when lightis yppgmixed, rather than when pigments are mixed using subtractive color theory.Color TheoryAdditive ColorSubtractive ColorIn additive color theory a pixel having RGB values of 255, 255, 255 produces a bright white pixel. Conversely, we would get a dark pigment if we mixed equally high proportions of blue, green, and red paint (subtractive color theory). Usingof blue, green, and red paint (subtractive color theory). Using three 8-bit images and additive color theory, we can display 16,777,216 color combinations.RGB Color Coordinate SystemRGB brightness values of 255RGB brightness values of 2552552550 yield a bright yellow pixel0 yield a bright yellow pixelRGB brightness values of 255,RGB brightness values of 255,255,255,0 yield a bright yellow pixel 0 yield a bright yellow pixel RGB brightness values of 255,RGB brightness values of 255, 0,0, 0 yield a bright red pixel0 yield a bright red pixelRGB values of 0,RGB values of 0, 0,0, 0 yield a black pixel0 yield a black pixelGrays are produced along the Grays are produced along the gray line in the RGB color gray line in the RGB color gygycoordinate system when equal coordinate system when equal proportions of blue, green, proportions of blue, green, and red are encountered (e gand red are encountered (e gand red are encountered (e.g., and red are encountered (e.g., an RGB of 127,an RGB of 127, 127,127, 127)127)88--bit Color Lookbit Color Look--up Tablesup TablesTheThe graygray tonetone oror colorcolor ofof anan individualindividual pixelpixel onon aa computercomputer screenscreenisis controlledcontrolled byby thethe sizesize andand characteristicscharacteristics ofof aa separateseparate blockblock ofofcomputercomputer memorymemory calledcalled aa colorcolor looklook--upup tabletable..ContainsContains thethe exactexact dispositiondisposition ofof eacheach combinationcombination ofof red,red, green,green,andand blueblue valuesvalues associatedassociated withwith eacheach 88--bitbit pixelpixel.. EvaluatingEvaluating thethenaturenature ofof anan 88--bitbit imageimageprocessorprocessor andandassociatedassociatedcolorcolor looklook--upuptabletable providesprovides insightinsight intointo thethe wayway thethe remoteremote sensingsensing brightnessbrightnessvaluesvalues andandcolorcolor looklook--upuptabletable interactinteract..ppThink about contrastColor Density Slice of the Thermal Infrared Image of the Savannah RiverThermal Infrared Image of the Savannah River24-bit Digital ImageImage Processing SystemSystemColor Composite of Marco Island, Florida SPOT Imagery October 11, 1988R = SPOT band 3 (NIR) 20 mG = SPOT band 4 (Pan) 10 mG = SPOT band 4 (Pan) 10 mB = SPOT band 1 (Green) 20 mColor CompositingOptimum Index FactorpRanks the 20 three-band combinations that can be made from six reflective bands of Landsat TM data (not including the thermal-infrared band).Band combination:∑==31kksOIFBand combination:1,2,31,2,41,2,5()∑==31jjrAbsOIF1,2,62,3,42,3,52,3,6jWhere skis the standard deviation for band k, and rjis the absolute value of the 2,3,63,4,53,4,6 etc.k,jcorrelation coefficient between any two of the three bands being evaluated. The largest OIF will generally have the most information (as measured by variance) with the least amount of duplication (as measured by correlation). Applicable to any multispectral dataset.Optimum Index FactorOptimum Index FactorRanks the 20 three-band combinations that can be made from six bands of Landsat TM data∑3ks()∑∑==31jkkrAbsOIF7539333 ++OIF69333++OIF()∑=1jjBand combination:1,2,3Band