Created
October 30, 2012 15:48
-
-
Save dsparks/3981069 to your computer and use it in GitHub Desktop.
Image Manipulation, Part 4
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
doInstall <- TRUE # Change to FALSE if you don't want packages installed. | |
toInstall <- c("ReadImages", "reshape", "ggplot2") | |
if(doInstall){install.packages(toInstall, repos = "http://cran.r-project.org")} | |
lapply(toInstall, library, character.only = TRUE) | |
imageLoader <- function(url){ # This function takes a URL, and generates a data.frame with pixel locations and colors | |
# Download to disk, load | |
download.file(url, "tempPicture.jpg", mode = "wb") # Stash image locally | |
readImage <- read.jpeg("tempPicture.jpg") | |
longImage <- melt(readImage) | |
rgbImage <- reshape(longImage, timevar = "X3", | |
idvar = c("X1", "X2"), direction = "wide") | |
rgbImage$X1 <- -rgbImage$X1 | |
return(rgbImage) | |
} | |
########## | |
# Part 4 # Finding a bright object | |
########## | |
# Load NASA's image | |
rgbImage <- imageLoader("http://www.nasa.gov/images/content/694811main_pia16225-43_full.jpg") | |
with(rgbImage, plot(X2, X1, col = rgb(rgbImage[, 3:5]), asp = 1, pch = ".")) | |
# Look at a region of only sand | |
sandOnly <- with(rgbImage, rgbImage[X2 > 1000 & X1 > -300, ]) | |
with(sandOnly, plot(X2, X1, col = rgb(sandOnly[, 3:5]), asp = 1, pch = ".")) | |
sandMean <- apply(sandOnly[, 3:5], 2, mean) # Find typical sand color values | |
sandSD <- apply(sandOnly[, 3:5], 2, sd) | |
colorZ <- sweep(rgbImage[, 3:5], 2, sandMean, "-") # How much do all colors | |
colorZ <- sweep(colorZ, 2, sandSD, "/") # deviate from sand colors? | |
plot(density(rowSums(colorZ))) | |
# Plot a "binarized" image, based on color Z-scores | |
with(rgbImage, plot(X2, X1, col = rgb(colorZ>4), asp = 1, pch = ".")) | |
points(830, -920, cex = 4, col = "RED") | |
# That little blue dot, at about 830, -920, is the anomaly! Science! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I found your code as a sample tutorial in stackoverflow, but I found this code is kind of out of data.
The
ReadImages
package is removed by cran, and substituted by thejpeg
package. And the function is nowreadJPEG()
.The result for the image loading might be a little bit different.
This is really a good example for image recognition, if it is okay, hope that you could update this code.
Thank you!