amannirala13/haar-cascade-trainer-linux

🖼️ This is a HAAR Cascade Classifier training GUI application for Linux. This application make it really easy to train classifiers for object detection and tracking using opencv by providing a Graphical user interface to set parameters and perform necessary steps.

haar-cascade-trainer-linux's Issues

libopencv_calib3d.so.3.2 and libopencv_imgcodecs.so.3.2 not found

Starting...
git version 2.25.1
opencv_createsamples: error while loading shared libraries: libopencv_calib3d.so.3.2: cannot open shared object file: No such file or directory
Environment ready!
Positive Dir selected: /home/sachin/Documents/Profession/OBJECTS/Face/p
Negative Dir selected: /home/sachin/Documents/Profession/OBJECTS/Face/n
Negative Dir selected:
Negative Dir selected: /home/sachin/Documents/Profession/OBJECTS/Face/n
Output Dir selected: /home/sachin/Documents/Profession/OBJECTS/Face
Successfully generated negative images index
Successfully generated positive images index
opencv_createsamples: error while loading shared libraries: libopencv_calib3d.so.3.2: cannot open shared object file: No such file or directory
mkdir: cannot create directory ‘classifier’: File exists
opencv_traincascade: error while loading shared libraries: libopencv_imgcodecs.so.3.2: cannot open shared object file: No such file or directory
Training Completed Successfully

Trained classifier detects too many false positives and usage percent not working

The classifiers trained detects a lot of false positives. The positive.vec file is generated successfully without any issues so probably the issue is with the test data set or the training parameters.

The usage percent option throws cv:Exception that insufficient number is provided than what is present in the positive.vec file.

The training is done. But it is not generating the XML file in the output directory

The training is fine. It created a new directory called classifier and also the vector file. But the classifier directory is empty. It is not generating the XML file. What to do?

opencv_createsamples: command not found

Hi, the contributor and others. I am having the same problem as described previously by others.
opencv_createsamples: command not found
My OpenCV ver is 4.5 and python is 3.10

would be really nice to see any updates on this by the contributor. Thanks, we appreciate your contribution

It's not working, Please make one video for this

Starting...
git version 2.25.1
sh: 1: opencv_createsamples: not found
Environment ready!
Positive Dir selected: /home/sachin/Documents/Profession/OBJECTS/Face/p
Negative Dir selected: /home/sachin/Documents/Profession/OBJECTS/Face/n
Output Dir selected: /home/sachin/Documents/Profession/OBJECTS/Face
Successfully generated negative images index
Successfully generated positive images index
sh: 1: opencv_createsamples: not found
ERROR: Couldnt train network for some unknown error!

trainer is not working

Nice job for making the trainer!

I am getting an error:
Environment ready!
objc[40735]: Class FIFinderSyncExtensionHost is implemented in both /System/Library/PrivateFrameworks/FinderKit.framework/Versions/A/FinderKit (0x7fff94ee23d8) and /System/Library/PrivateFrameworks/FileProvider.framework/OverrideBundles/FinderSyncCollaborationFileProviderOverride.bundle/Contents/MacOS/FinderSyncCollaborationFileProviderOverride (0x122b4bf50). One of the two will be used. Which one is undefined.
Positive Dir selected: /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Positive
Negative Dir selected: /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Negative
Positive Dir selected: /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Positive
Negative Dir selected: /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Negative
(base) Nishchays-MacBook-Pro:app nishchay$ chmod -R a+x ./app
chmod: ./app: No such file or directory
(base) Nishchays-MacBook-Pro:app nishchay$ python3 main.py
Starting...
git version 2.28.0
sh: opencv_createsamples: command not found
Environment ready!
objc[40763]: Class FIFinderSyncExtensionHost is implemented in both /System/Library/PrivateFrameworks/FinderKit.framework/Versions/A/FinderKit (0x7fff94ee23d8) and /System/Library/PrivateFrameworks/FileProvider.framework/OverrideBundles/FinderSyncCollaborationFileProviderOverride.bundle/Contents/MacOS/FinderSyncCollaborationFileProviderOverride (0x128c25f50). One of the two will be used. Which one is undefined.
Positive Dir selected: /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Positive
Negative Dir selected: /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Negative
Output Dir selected: /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training
sh: -c: line 0: syntax error near unexpected token (' sh: -c: line 0: find /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Negative -name '.png' -o -name '.jpg' > /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Negative/index.txt'
Successfully generated negative images index
sh: -c: line 0: syntax error near unexpected token (' sh: -c: line 0: find /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Positive -name '.png' -o -name '.jpg' > /Users/nishchay/Dropbox/My Mac (Nishchay’s MacBook Pro)/Desktop/Training/Positive/index.txt'
Successfully generated positive images index
ERROR: Unable to open positive images index file.

Are the required dimensions related to the postive or negative images?

What about the width and height

• Are they related to the positive or negative images?
• Negative and positive image should have the same dimensions?

Also in readme there is a typo: traget instead target
Now provide the width and height of the traget object which is by default set to 24x24.

Thanks

sh: 1: opencv_createsamples: not found sh: 1: ./dir_gen.sh: not found ERROR: Couldnt train network for some unknown error!

info:
linux parrot os 4.11.2 debian KDE plasma.
Python 3.9.2
Git, Tkinter, OpenCV 3.2.0, LibOpenCV
all installed, and working.

Created files before running trainer:
282 photos of differnt positions of the same face, with 99% white background.
100 negative photos differnt body parts and pictures that are not a face.

App settings:
sample image: W = 24
sample image: H = 24
sample image usage percent(%) = 100
number of traning stages 10

error:
sh: 1: opencv_createsamples: not found
sh: 1: ./dir_gen.sh: not found
ERROR: Couldnt train network for some unknown error!

Everything working before this error happened.

Unable to run HAAR-Cascade-Trainer-GUI on Ubuntu 20.04

Hi Team,

I was trying to Run HAAR Cascade GUI Trainer 1.0 on my Ubuntu 20.04,
I have followed the Requirement you mentioned.

When I was trying to run the app/main.py getting an error.

Starting...
git version 2.25.1
sh: 1: opencv_createsamples: not found
Environment ready!
No protocol specified
No protocol specified
Traceback (most recent call last):
File "main.py", line 187, in
main_window = tk.Tk()
File "/usr/lib/python3.8/tkinter/init.py", line 2270, in init
self.tk = _tkinter.create(screenName, baseName, className, interactive, wantobjects, useTk, sync, use)
_tkinter.TclError: couldn't connect to display ":0"

can you please help me, to run the Trainer application?.