Western+Brazil+Landsat

__1999__
 * Lab 11 & Lab 12**







__2007__







__Change Detection__




 * Lab 10**












 * Lab 9**

Original Subset of the image - from July 29, 1999

Sobel Edge Detector - For this image I used the Sobel filter to sharpen the edges of the image. This filter really brought out the edges that are found in the image and highlighted them so that it is easy to see the changes from forest to deforested.

Median Filter - For this filter I used a Median Filter with a kernel size of 3. This filter did the almost the opposite of the sobel filter. It smoothed everything out. When I zoomed in the image looked much less pixelated.

Median Filter - For this filter I used the same filter as before but with a much larger kernel size of 13. This smoothed the image even more than the 3 kernel filter. The difference was even visible in the large image. It have almost no distinct lines, all of the lines sort of flow into the others.

In the PCs I can see that a lot of them are very noisy. The first one looks very similar to the original image, but they get worse from there. The second one appears to contain barely anything useful but when you zoom in the deforested areas come out very clearly.

Basic Stats Min Max Mean Stdev Num Eigenvalue Band 1 -1389 5991 185.677357 120.730525 1 761987.471342 Band 2 -58 2201 419.925139 202.131874 2 185183.158808 Band 3 -414 6227 403.972648 290.961357 3 11046.111967 Band 4 -92 6083 2832.883395 427.508099 4 5050.373641 Band 5 -10 7077 1755.803815 642.115051 5 1567.604459 Band 6 -27 9422 833.267966 480.647357 6 1353.739100

Covariance Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 1 14575.859577 22657.259819 32878.235258 3425.824959 69938.182158 52066.162981 Band 2 22657.259819 40857.294476 56483.870782 14616.127591 119746.045353 87571.471353 Band 3 32878.235258 56483.870782 84658.511453 3661.445427 176485.404244 132889.401532 Band 4 3425.824959 14616.127591 3661.445427 182763.174290 29176.091785 -5719.116168 Band 5 69938.182158 119746.045353 176485.404244 29176.091785 412311.738113 300497.630927 Band 6 52066.162981 87571.471353 132889.401532 -5719.116168 300497.630927 231021.881408

Correlation Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 1 1.000000 0.928444 0.935957 0.066375 0.902162 0.897247 Band 2 0.928444 1.000000 0.960405 0.169143 0.922600 0.901366 Band 3 0.935957 0.960405 1.000000 0.029436 0.944628 0.950229 Band 4 0.066375 0.169143 0.029436 1.000000 0.106284 -0.027833 Band 5 0.902162 0.922600 0.944628 0.106284 1.000000 0.973647 Band 6 0.897247 0.901366 0.950229 -0.027833 0.973647 1.000000

Quite a few of the bands have correlations higher than 0.9. Band 5 and Band 6 have the highest correlation with a 0.97. Band 2 and 3 are also highly correlated with a value of 0.96. Band 3 and Band 6 have a correlation value of 0.95. The lowest correlation was a -0.027 between Band 4 and 6, but Band 3 and 4 also had a low correlation value of 0.029. I am confused as to why 4 and 6 are negatively correlated. They are the only two bands that are like this and I'm not sure if this is supposed to happen considering everything else is positively correlated.

Eigenvector Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 1 0.127298 0.217576 0.321906 0.039808 0.732393 0.542999 Band 2 0.002110 -0.044708 0.033397 -0.989914 -0.039853 0.123946 Band 3 0.280880 0.554393 0.658611 -0.003122 -0.401722 -0.136366 Band 4 -0.069427 -0.184956 0.019980 0.131479 -0.544955 0.803934 Band 5 0.726724 0.325395 -0.590496 -0.015972 -0.051527 0.119980 Band 6 0.609877 -0.709379 0.335274 0.030552 0.031644 -0.102413

Over 99% of the variability is explained in the first three PCs. 78.8% is explained in just the first PC, with 19.2% and 1.1% explained by PCs 2 and 3 respectively.
 * PC || Eigenvalue ||  || % explained ||
 * 1 || 761987.5 ||  || 0.788653 ||
 * 2 || 185183.2 ||  || 0.191664 ||
 * 3 || 11046.11 ||  || 0.011433 ||
 * 4 || 5050.374 ||  || 0.005227 ||
 * 5 || 1567.604 ||  || 0.001622 ||
 * 6 || 1353.739 ||  || 0.001401 ||
 * 6 || 1353.739 ||  || 0.001401 ||

I took the first three PCs and made an inverse PC rotation image. The image differed a little from the original image. It appears much clearer and brighter. It is much easier to look at it and interpret. The forests are much more defined, its possible to see individual trees. The colors in the image are more uniform, unlike in the first image where it almost looks like there are thin clouds or different lights shining on different areas.




 * Lab 7 |||| Amanda Hoffman ||  ||   ||
 * Path || 231 ||  ||   ||   ||
 * Row || 68 ||  ||   ||   ||
 * Available Scenes ||  ||   ||   ||   ||
 * ID || Date || Cloud Cover || Quality || Downloadable ||
 * LT52310682009221CUB00 || 8/9/2009 || 0% || 9 || Yes ||
 * LT52310682008219CUB00 || 8/6/2008 || 8% || 9 || No ||
 * LT52310682008203CUB00 || 7/21/2008 || 0% || 7 || Yes ||
 * LT52310682007216CUB00 || 8/4/2007 || 0% || 9 || Yes ||
 * LT52310682007184CUB00 || 7/3/2007 || 0% || 9 || No ||
 * LT52310682006245CUB00 || 9/2/2006 || 6% || 9 || No ||
 * LT52310682006197CUB00 || 7/16/2006 || 0% || 7 || No ||
 * LT52310682006133CUB00 || 5/13/2006 || 0% || 7 || No ||
 * LT52310682005274CUB00 || 10/1/2005 || 2% || 9 || No ||
 * LT52310682005210CUB00 || 7/29/2005 || 0% || 9 || No ||
 * LT52310682005194CUB00 || 7/13/2005 || 0% || 9 || Yes ||
 * LT52310682005162CUB00 || 6/11/2006 || 0% || 9 || No ||
 * LT52310682004224CUB00 || 8/11/2004 || 0% || 9 || Yes ||
 * LT52310682003205CUB00 || 7/24/2003 || 0% || 9 || Yes ||
 * LT52310682001215CUB00 || 8/3/2001 || 0% || 9 || No ||
 * LT52310682000261XXX02 || 9/17/2000 || 0% || 9 || No ||
 * LT52310682000229XXX02 || 8/16/2000 || 0% || 9 || Yes ||
 * LT52310682000197XXX02 || 7/15/2000 || 0% || 9 || No ||
 * LT52310682000181XXX02 || 6/29/2000 || 0% || 9 || No ||
 * LT52310682000133XXX02 || 5/12/2000 || 0% || 9 || No ||
 * LT52310681999290XXX01 || 10/17/1999 || 0% || 9 || No ||
 * LT52310681999242XXX02 || 8/30/1999 || 0% || 9 || No ||
 * LT52310681999210CUB00 || 7/29/1999 || 0% || 9 || Yes ||
 * LT52310681999194CUB02 || 7/13/1999 || 2% || 9 || No ||
 * LT52310681999178XXX02 || 6/27/1999 || 10% || 9 || No ||
 * LT52310681998255XXX01 || 9/12/1998 || 0% || 9 || No ||
 * LT52310681998207XXX00 || 7/26/1998 || 0% || 9 || No ||
 * LT52310681998191XXX01 || 7/10/1998 || 0% || 9 || No ||
 * LT52310681998175XXX01 || 6/24/1998 || 20% || 9 || No ||
 * LT52310681998159XXX01 || 6/8/1998 || 0% || 9 || No ||
 * LT52310681998143XXX02 || 5/23/1998 || 0% || 9 || No ||
 * LT52310681997172XXX02 || 6/21/1997 || 0% || 9 || No ||
 * LT52310681997156XXX01 || 6/5/1997 || 0% || 9 || No ||
 * LT52310681996266XXX01 || 9/22/1996 || 0% || 9 || No ||
 * LT52310681996202XXX02 || 7/20/1996 || 0% || 9 || No ||
 * LT52310681996186XXX01 || 7/4/1996 || 0% || 9 || No ||
 * LT52310681995215CUB00 || 8/3/1995 || 0% || 7 || Yes ||
 * LT52310681994228CUB00 || 8/16/1994 || 2% || 9 || No ||
 * LT52310681994212CUB02 || 7/31/1994 || 0% || 9 || No ||
 * LT52310681993209CUB02 || 7/28/1993 || 12% || 9 || No ||
 * LT52310681992207CUB00 || 7/25/1992 || 0% || 9 || Yes ||
 * LT52310681991268CUB00 || 9/25/1991 || 0% || 9 || Yes ||
 * LT52310681991220CUB02 || 8/8/1991 || 1% || 9 || No ||
 * LT52310681990169CUB00 || 6/18/1990 || 0% || 9 || No ||
 * LT52310681989214CUB02 || 8/2/1989 || 0% || 7 || Yes ||
 * LT52310681989198CUB00 || 7/17/1989 || 0% || 9 || Yes ||
 * LT52310681987193CUB02 || 7/12/1987 || 0% || 9 || No ||
 * LT52310681987001XXX03 || 1/1/1987 || 10% || 9 || No ||
 * LT52310681986286XXX02 || 10/13/1986 || 0% || 9 || No ||
 * LT52310681986254XXX03 || 9/11/1986 || 20% || 9 || No ||
 * LT52310681986222XXX03 || 8/10/1986 || 0% || 9 || Yes ||
 * LT52310681986206XXX03 || 7/25/1986 || 0% || 7 || No ||
 * LT52310681986110XXX03 || 4/20/1986 || 20% || 7 || No ||
 * LT52310681986206XXX03 || 7/25/1986 || 0% || 7 || No ||
 * LT52310681986110XXX03 || 4/20/1986 || 20% || 7 || No ||

July 29, 1999

August 4th, 2007