Hi!
I tired this code:

yolo_tiny_voc <- image_darknet_model(type = 'detect', model = "tiny-yolo-voc.cfg", weights=system.file(package="image.darknet", "models", "tiny-yolo-voc.weights"), labels = system.file(package="image.darknet", "include", "darknet", "data", "voc.names"))

x<-image_darknet_detect(paste(getwd(), "/tempPicture.jpg", sep=""), object = yolo_tiny_voc)

The output of this code is :

layer     filters    size              input                output

    0 conv     16  3 x 3 / 1   416 x 416 x   3   ->   416 x 416 x  16

    1 max          2 x 2 / 2   416 x 416 x  16   ->   208 x 208 x  16

    2 conv     32  3 x 3 / 1   208 x 208 x  16   ->   208 x 208 x  32

    3 max          2 x 2 / 2   208 x 208 x  32   ->   104 x 104 x  32

    4 conv     64  3 x 3 / 1   104 x 104 x  32   ->   104 x 104 x  64

    5 max          2 x 2 / 2   104 x 104 x  64   ->    52 x  52 x  64

    6 conv    128  3 x 3 / 1    52 x  52 x  64   ->    52 x  52 x 128

    7 max          2 x 2 / 2    52 x  52 x 128   ->    26 x  26 x 128

    8 conv    256  3 x 3 / 1    26 x  26 x 128   ->    26 x  26 x 256

    9 max          2 x 2 / 2    26 x  26 x 256   ->    13 x  13 x 256

   10 conv    512  3 x 3 / 1    13 x  13 x 256   ->    13 x  13 x 512

   11 max          2 x 2 / 1    13 x  13 x 512   ->    13 x  13 x 512

   12 conv   1024  3 x 3 / 1    13 x  13 x 512   ->    13 x  13 x1024

   13 conv   1024  3 x 3 / 1    13 x  13 x1024   ->    13 x  13 x1024

   14 conv    125  1 x 1 / 1    13 x  13 x1024   ->    13 x  13 x 125

   15 detection

Loading weights from   ---------------------/R/x86_64-pc-linux-gnu-library/3.3/image.darknet/models/tiny-yolo-voc.weights...Done!

--------------------------/tempPicture.jpg: Predicted in 2.397771 seconds.
Boxes: 845 of which 1 above the threshold.

person: 93%

So the model detect one person into the image.
I wish save into x that te model detect one person. But the value of x is NULL

I tried also the function capture.output() to capture the fact that the model find a person in the image but the result of

capture.output(image_darknet_detect(paste(getwd(), "/tempPicture.jpg", sep=""), object=yolo_tiny_voc))

is charather(0).

Could you help me? It seem impossible extract the result of the detection.

Ty

P.S. I would detect 1000 image and put the result of the detection in a model like randomForset in order to predict a value for this reason I wish put the result of the detection in a data.frame

Hello,
Is it possible to get the "file" for image_darknet_detect from any memory object like array/mtx etc...
This can be a cool idea for processing images from webcam/cam stream without dumping them to a file and reading them back from the file.
Thanks for your effort...

Similarly to #7 and #19, when I set union = T in image_line_segment_detector(), R crashes. The following reproduces the error for me:

library(image.LineSegmentDetector)
library(pixmap)

image <- read.pnm(file = system.file("extdata", "le-piree.pgm", package="image.LineSegmentDetector"), cellres = 1)

linesegments <- image_line_segment_detector(image@grey * 255, union = T)

R version 4.0.5 (2021-03-31)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur 10.16

Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib

locale:
[1] en_GB.UTF-8/en_GB.UTF-8/en_GB.UTF-8/C/en_GB.UTF-8/en_GB.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] pixmap_0.4-12                   image.LineSegmentDetector_0.1.0

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.7         pillar_1.6.3       compiler_4.0.5     class_7.3-19       tools_4.0.5       
 [6] digest_0.6.28      jsonlite_1.7.2     evaluate_0.14      lifecycle_1.0.1    tibble_3.1.5      
[11] lattice_0.20-44    pkgconfig_2.0.3    rlang_0.4.11       DBI_1.1.1          yaml_2.2.1        
[16] xfun_0.26          fastmap_1.1.0      e1071_1.7-9        dplyr_1.0.7        httr_1.4.2        
[21] knitr_1.36         raster_3.4-13      generics_0.1.0     vctrs_0.3.8        classInt_0.4-3    
[26] grid_4.0.5         tidyselect_1.1.1   glue_1.4.2         sf_1.0-2           R6_2.5.1          
[31] fansi_0.5.0        rmarkdown_2.10     bookdown_0.23      sp_1.4-5           purrr_0.3.4       
[36] clipr_0.7.1        magrittr_2.0.1     units_0.7-2        codetools_0.2-18   ellipsis_0.3.2    
[41] htmltools_0.5.2    assertthat_0.2.1   KernSmooth_2.23-20 utf8_1.2.2         proxy_0.4-26      
[46] crayon_1.4.1 

For example, in this picture of people in Lima, Peru, it doesn't detect a face: https://www.maxpixel.net/static/photo/1x/South-America-Lima-Peru-Road-Colorful-Color-1244329.jpg

Another example, this time from Wikimedia, it detects 3 faces: https://upload.wikimedia.org/wikipedia/commons/e/ec/Urarina_shaman_B_Dean.jpg

Using a smaller resolution from the second image, it doesn't detect any face (source: https://upload.wikimedia.org/wikipedia/commons/thumb/e/ec/Urarina_shaman_B_Dean.jpg/406px-Urarina_shaman_B_Dean.jpg)

I've used exactly the same code as the article in http://www.bnosac.be/index.php/blog/89-human-face-detection-with-r

library(magick)
library(image.libfacedetection)

# using an image where it fails to detect a face
img_url <- "https://upload.wikimedia.org/wikipedia/commons/thumb/e/ec/Urarina_shaman_B_Dean.jpg/406px-Urarina_shaman_B_Dean.jpg"

image <- image_read(img_url)
faces <- image_detect_faces(image)
faces
plot(faces, image, border = "red", lwd = 7, col = "white")

I am showing below the results from printing the faces object:

> faces
$nr
[1] 0

$detections
[1] x          y          width      height     neighbours
[6] angle     
<0 rows> (or 0-length row.names)

attr(,"class")
[1] "libfacedetection"

Is it a problem with the image resolution perhaps? If so, then it should be noted in the documentation, and perhaps change the comment in https://github.com/bnosac/image/blob/master/image.libfacedetection/man/image_detect_faces.Rd which in line 20 indicates:

The function can be used to detect faces of minimal size 48x48 pixels.

Reproducibility information

> R.version
               _                           
platform       x86_64-w64-mingw32          
arch           x86_64                      
os             mingw32                     
system         x86_64, mingw32             
status                                     
major          3                           
minor          5.2                         
year           2018                        
month          12                          
day            20                          
svn rev        75870                       
language       R                           
version.string R version 3.5.2 (2018-12-20)
nickname       Eggshell Igloo
> packageVersion("magick")
[1] ‘2.0’
> packageVersion("image.libfacedetection")
[1] ‘0.1’

Putting here compilation issues at CRAN machines

Hi!
I tried to make detection with this two models: yolo_tiny_voc and yolo_tiny_coco. The first model detect the object but the second one don't detect anything. The code I used is it:

yolo_tiny_voc <- image_darknet_model(type = 'detect', 
                                     model = "tiny-yolo-voc.cfg", 
                                     weights = system.file(package="image.darknet", "models", "tiny-yolo-voc.weights"), 
                                     labels = system.file(package="image.darknet", "include", "darknet", "data", "voc.names"))
yolo_tiny_voc

x <- image_darknet_detect("__path__", object = yolo_tiny_voc)

###################################
yolo_tiny_coco <- image_darknet_model(type = 'detect', 
                                     model = "tiny-yolo-voc.cfg", 
                                     weights = system.file(package="image.darknet", "models", "tiny-yolo.weights"),
                                     labels = system.file(package="image.darknet", "include", "darknet", "data", "coco.names"))

##
x <- image_darknet_detect("__path__", object = yolo_tiny_coco)

I would find a model with 80 labels that works in a detection model. I found only a model with 20 labels that works for detection.

Thank you

Using image.draknet pkg on R, with original trained Yolo weights, the "labels" class list length in the function image_darknet_model must be larger than 1?
but I trained only 1 class. how can i do??

Hi,

Trying to install image.ContourDetector on windows through rstudio, I get this error.
Did I miss something ?

Thanks.

RcppExports.cpp:24:58: error: 'image' was not declared in this scope
{"image.ContourDetector_detect_contours", (DL_FUNC) &image.ContourDetector_detect_contours, 4},
^
RcppExports.cpp:28:29: error: expected initializer before '.' token
RcppExport void R_init_image.ContourDetector(DllInfo *dll) {
^
make: *** [RcppExports.o] Error 1
Warning: l'exécution de la commande 'make -f "C:/PROGRA1/R/R-331.2/etc/x64/Makeconf" -f "C:/PROGRA1/R/R-331.2/share/make/winshlib.mk" SHLIB_LDFLAGS='$(SHLIB_CXXLDFLAGS)' SHLIB_LD='$(SHLIB_CXXLD)' SHLIB="image.ContourDetector.dll" WIN=64 TCLBIN=64 OBJECTS="RcppExports.o contour_detector.o image.ContourDetector_init.o smooth_contours.o"' renvoie un statut 2
ERROR: compilation failed for package 'image.ContourDetector'

image_line_segment_detector(img, union = T) returns a pixel matrix the size of the image with all 0 values. Ideally, this would also update when union = T so that joined lines are represented by the same number.

library(imager)
library(image.LineSegmentDetector)

im <- load.image(system.file('extdata/parrots.png', package='imager'))

im_pnm <- im %>%
  grayscale() %>%
  multiply_by(255)
im_pnm <- im_pnm[ , , 1, 1]

nounion <- image_line_segment_detector(im_pnm)
withunion <- image_line_segment_detector(im_pnm, union = T)

sum(nounion$pixels != 0) # 19414
sum(withunion$pixels != 0) # 0

Hi there,

I am trying to print the output of the function image_darknet_detect, but I'm unable to figure it out. More specifically: I'm trying to print the bounding boxes and the labels of the boxes (e.g. persons: 22%, car: 89%). Can anybody tell me how to do that?

I tried to get this information out of the object, but it didn't work.
Help is appreciated.

I've like to use the image_darknet_model() with my custom YOLOv3 model. For this, first I put my obj_100.weights file that represents the weights of my custom model create using darknet53.conv.74 inside the library directory image.darknet/models. After, I make the same with my obj.cfg in image.darknet/include/darknet/cfg directory and obj.names in image.darknet/include/darknet/data directory. Then, I try to use the image_darknet_model() function:

library(image.darknet)
yolo_v3_my_weights <- image_darknet_model(type = 'classify',
                                           model = "obj.cfg", weights = system.file(package="image.darknet", "models", "obj_100.weights"), 
labels = system.file(package="image.darknet", "include", "darknet", "data", "obj.names"))
Error in image_darknet_model(type = "classify", model = "obj.cfg", weights = system.file(package = "image.darknet",  : 
  file.exists(model) is not TRUE

But the I have the model in correct directory:

setwd("~/R/win-library/3.6/image.darknet/models")
dir()
[1] "obj_100.weights"       "tiny-yolo-voc.weights" "tiny.weights"

The file is OK and I don't have any problem with obj_100.weights file in the darknet in Python.

Please, any solution for this?

I have been trying to use the image.darknet library as part of a looping process to process multiple files. It appears that doing so causes a crash of the R session. It happens using a fresh download of the library on multiple computers. On each computer, the crash consistantly happens, but happens at a different number of images. For example, on one computer the R session crashes on processing the fourth file. On another, it is always after the 18th file. This happens if the detect is called as part of a loop, a foreach package loop, or just a run-on of detect calls in the script. Here is a reproducible example that works on any computer I seem to run it on.

`
devtools::install_github("bnosac/image", subdir = "image.darknet", build_vignettes = TRUE)
library(image.darknet)

labels<-system.file(package="image.darknet", "include", "darknet", "data", "voc.names")
labels<-readLines(labels)
yolo_tiny_voc<-image_darknet_model(
type="detect",
model="tiny-yolo-voc.cfg",
weights=system.file(package="image.darknet", "models", "tiny-yolo-voc.weights"),
labels=labels)

f <- system.file("include", "darknet", "data", "dog.jpg", package="image.darknet")

for (i in 1:50){
print(i)
image_darknet_detect(file=f, object=yolo_tiny_voc, threshold=0.03)
}
`

Hi,
I get an error trying to install darknet.

devtools::install_github("bnosac/image", subdir = "image.darknet", build_vignettes = TRUE)

```detector.o:detector.c:(.text+0x17cb):undefined reference topthread_join'
detector.o:detector.c:(.text+0x1844): undefined reference to `pthread_join'
detector.o:detector.c:(.text+0x18c1): more undefined references to `pthread_join' follow
collect2: ld returned 1 exit status
keine DLL erzeugt
ERROR: compilation failed for package 'image.darknet'

• removing 'C:/Users/fteschner/Documents/R/win-library/3.2/image.darknet'
  Error: Command failed (1)```

R version 3.2.4 Revised (2016-03-16 r70336)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1

locale:
[1] LC_COLLATE=German_Germany.1252  LC_CTYPE=German_Germany.1252    LC_MONETARY=German_Germany.1252
[4] LC_NUMERIC=C                    LC_TIME=German_Germany.1252    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] dlib_1.0

loaded via a namespace (and not attached):
 [1] httr_1.2.1      R6_2.2.0        tools_3.2.4     withr_1.0.2     curl_2.5        memoise_1.1.0   Rcpp_0.12.10   
 [8] git2r_0.18.0    digest_0.6.12   devtools_1.12.0```

I don't find any material about object detection using *h5 cnn adjusted model.
I try to create a simple example because I don't find in any book/web, I've like to apply any object detection technique to find handwritten digits in an image with several digits (wh_img). First, a create a CNN using the classic MNIST example:

Preparing the Data - using MNIST dataset is included with Kera

library(keras)
mnist <- dataset_mnist()
x_train <- mnist$train$x
y_train <- mnist$train$y
x_test <- mnist$test$x
y_test <- mnist$test$y

#reshape
x_train <- array_reshape(x_train, c(nrow(x_train), 784))
x_test <- array_reshape(x_test, c(nrow(x_test), 784))
#rescale
x_train <- x_train / 255
x_test <- x_test / 255

y_train <- to_categorical(y_train, 10)
y_test <- to_categorical(y_test, 10)

CNN model

model <- keras_model_sequential()
model %>%
layer_dense(units = 256, activation = 'relu', input_shape = c(784)) %>%
layer_dropout(rate = 0.4) %>%
layer_dense(units = 128, activation = 'relu') %>%
layer_dropout(rate = 0.3) %>%
layer_dense(units = 10, activation = 'softmax')

model %>% compile(
loss = 'categorical_crossentropy',
optimizer = optimizer_rmsprop(),
metrics = c('accuracy')
)

#Training and Evaluation
history <- model %>% fit(
x_train, y_train,
epochs = 30, batch_size = 128,
validation_split = 0.2
)
plot(history)

Save the model

model %>% save_model_hdf5("cnn_digits.h5")

Now, I've to apply any object detection technique in my y.png image with several handwritten digits to find only 4 numerals using my trained model :

Open an image with handwrite numbers collections

y = "http://mariakravtsova.us/img/numbers.png"
download.file(y,'y.png', mode = 'wb')
library("png")
wh_img <- readPNG("y.png")
plot.new()
rasterImage(wh_img,0,0,1,1)

library(image.darknet)

Try to use *h5 model to object detection only 4 handwritten digits

yolo_digits <- image_darknet_model(type = ‘detect’, labels ="4", model = "cnn_digits.h5")

x <- image_darknet_detect(file = "/Documents/numbers.png",
object = yolo_digits,
threshold = 0.4)

This is possible in R? Because in Python there are a lot of functions, choose the window size, etc. but in R not :(

Hi,

When running the image_darknet_detect function my R sessions crashes .
Do you have a solution for this problem?

Many thanks!

All images I have tried image_surf on return a surf vector of 64 dimension, but only the first dimension has a non-zero value. The function runs without warning or errors, and I have successfully used the exact same images to run the image_fhog function in the same package

This is true as well for the example code you give in the "computer vision for R users" presentation:

f <- system.file("extdata", "cruise_boat.bmp", package="image.dlib")
surf_blobs <- image_surf(f, max_points = 10000, detection_threshold = 50)

e.g., on the above example, from skimr
skim(as.data.frame(surf_blobs$surf))

Variable type: numeric
variable missing complete n mean sd p0 p25 median p75 p100 hist
V1 0 296 296 0.018 0.055 -0.089 0 0 0 0.32 ▁▇▁▁▁▁▁▁
V10 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V11 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V12 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V13 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V14 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V15 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V16 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V17 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V18 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V19 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V2 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V20 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V21 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V22 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V23 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V24 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V25 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V26 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V27 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V28 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V29 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V3 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V30 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V31 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V32 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V33 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V34 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V35 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V36 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V37 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V38 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V39 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V4 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V40 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V41 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V42 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V43 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V44 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V45 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V46 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V47 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V48 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V49 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V5 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V50 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V51 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V52 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V53 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V54 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V55 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V56 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V57 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V58 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V59 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V6 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V60 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V61 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V62 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V63 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V64 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V7 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V8 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁
V9 0 296 296 0 0 0 0 0 0 0 ▁▁▁▇▁▁▁▁

https://github.com/gfacciol/mgm
https://github.com/mnhrdt/imscript
https://github.com/kidanger/degradations-deconvolution
https://github.com/ShiqiYu/libfacedetection
https://cran.r-project.org/web/packages/CovTools/index.html
https://cran.r-project.org/web/packages/filling/filling.pdf
https://www.maths.nottingham.ac.uk/plp/pmzild/shapes/

https://github.com/dorian3d/DBoW2

https://github.com/lucasb-eyer/pydensecrf/tree/master/pydensecrf/densecrf
http://graphics.stanford.edu/projects/drf/

After having installed image.darknet in R I tried to carry out the model using the weights trained on "coco", just copying the code and nothing else. That said, I gave input a very simple "object" to detect (it was a sweetie cat), but the output has been the following:
C:////*/cat.jpg: Predicted in 5.22000 seconds.
Boxes: 845 of which 0 above the threshold.
After that I have tried with some other images whose content were bottles, dogs, so I guess the model shouldn't have any problem to detect them correctly since the coco model provides the possibility to detect those labels.
I suppose that I am missing something critical. Any help?
Thanks in advance for your answers!

Hi,
first of all: thanks for the implementation of such a tool!
I'm following the code in your example:

im <- image_read(file.path(out.folder,"raster","fence_0.25m.tif"))
im = image_data(im)
mat <- as.integer(im, transpose = TRUE)
mat <- drop(mat)



cld = image_contour_detector(mat)
plot(cld)

I'm using it to extract guard rails from an aerial image. It works pretty fine and super quickly even though all spatial references (coordinates) are lost in the process of conversion to matrix (but I think that can be solved with a workaround).

Right now I'm looking for a method to extract the lines from the cld object as a vector (sp or sf) object but I can't find any info in the data slot such as e.g. beginning of line, end of line, other. In addition, the x/y coords provided in the data slot do not correspond to cell positions (i.e. x = row number+1, y = to a decimal number; see example below).

      x        y curve
1     2 645.5518     1
2     3 645.0601     1
3     4 644.8012     1

Anything I'm missing? how are coords encoded?
Thanks in advance

• https://github.com/shorschig/Rcpp-CUDA-TestPackage/blob/master/src/Makevars for cuda + R + darknet
• replace image.darknet with code from https://github.com/AlexeyAB/darknet, factor out LSTM/GRU/RNN for text and for images (maybe create a different package for images / text)