Image processing 1

Question

Answered my question. Thanks a lot.

Answer 1

Here's one method.

Let's start by reading in your image. We'll use the png library for this, so if you don't have it installed, you'll want to install.packages("png") first.

library(png)

img_link <- "https://i.stack.imgur.com/PygvJ.png"

img <- readPNG(readBin(img_link, "raw", 1e6))

The readPNG function extracts the image as an array. In this case, the array is 360 by 371 by 4:

dim(img)
#> [1] 360 371   4

This means the image is 360 by 371 pixels and contains four channels: red, blue, green and alpha (transparency).

We can plot this array at any time by calling:

plot(as.raster(img))

To find a horizontal red line, all we need to do is look at the row sums of the pixel values in the green channel, which will drop sharply at the target row.

sum_green <- apply(img[,,2], 1, sum)

We can plot this vector to ensure there is a row with a large drop-off in one of the rows, as we would expect.

plot(sum_green, type = "l")

To get this row, we just find the minimum row sum:

target_row <- which.min(sum_green)

target_row
#> [1] 183

So our red line is on the 183rd row. To prove this, let's make the 183rd row of our image black and draw it again:

img[target_row, , 1:3] <- 0

plot(as.raster(img))

This looks correct.