get_grooves_hough: Use Hough transformation to identify groove locations.

In heike/grooveFinder: Identify Locations of Grooves in Land Engraved Areas of Bullet Scans

Description

Hough transformations are used to identify the location of left and right groove. Generates Hough lines and selects lines with angle corresponding to (close to) vertical lines.

Usage

get_grooves_hough(land, norm.index = 1, adjust = 10, return_plot = FALSE)

Arguments

land	Either a dataframe of surface measurements in microns in the x, y, and x direction, or an x3p file. Use 'x3p_to_df' to access the data from an x3p scan.
norm.index	positive number index to indicate which of the ordered normalized hough scores should be used to estimate the grooves area
adjust	(generally) positive number in micron used to adjust the grooves inward
return_plot	boolean value - should a plot of the crosscut with the grooves be returned? defaults to FALSE

Value

list object consisting of functions to describe the left and right groove. Parameters for the functions are given in microns and return results in microns.

Examples

library(x3ptools)
library(ggplot2)


x3p <- br411

# Get grooves fit for left and right groove from x3p
grooves <- get_grooves_hough(x3p_to_df(x3p))

# Find optimized crosscut location, this may take some time
# if (require(bulletxtrctr)) {
# crosscut <- x3p %>% bulletxtrctr::x3p_crosscut_optimize()
# } else {
crosscut <- 125
# }
## Not run: 
a <- get_mask_hough(x3p, grooves)
a <- a %>% x3p_add_hline(yintercept = crosscut)
x3ptools::image_x3p(a)

## End(Not run)

# Find groove locations for specified crosscut
grooves$left.groove.fit(crosscut)
grooves$right.groove.fit(crosscut)

# Plot profile
ccdata <- x3p %>%
  x3p_to_df() %>%
  dplyr::filter(y == crosscut)

ccdata %>%
  ggplot(aes(x = x, y = value)) +
  geom_line() +
  geom_vline(xintercept = grooves$left.groove.fit(crosscut), color = "red") +
  geom_vline(xintercept = grooves$right.groove.fit(crosscut), color = "blue")

# grooves at a different crosscut:
x3p %>%
  x3p_to_df() %>%
  dplyr::filter(y == 150) %>%
  ggplot(aes(x = x, y = value)) +
  geom_line() +
  geom_vline(xintercept = grooves$left.groove.fit(crosscut), color = "red") +
  geom_vline(xintercept = grooves$right.groove.fit(crosscut), color = "blue")

heike/grooveFinder documentation built on Dec. 31, 2019, 12:33 p.m.