puzzledqs / bbox-label-tool Goto Github PK
View Code? Open in Web Editor NEWA simple tool for labeling object bounding boxes in images
License: MIT License
A simple tool for labeling object bounding boxes in images
License: MIT License
Is it best to apply a bounding box to the largest, unobstructed view (which would be a similar approach to labeling objects at the edge of a picture where the complete object is not within the image) or is it best to apply a bounding box to the entire object (which could include some obstructions).
When I run the code on windows, I encounter the following error
////////////////////////////////////////////////////////
runfile('E:/Projects/Tensorflow/BBox-Label-Tool/main.py', wdir='E:/Projects/Tensorflow/BBox-Label-Tool')
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Program Files\Anaconda3\envs\tensorflow\lib\tkinter_init_.py", line 1559, in call
return self.func(*args)
File "E:/Projects/Tensorflow/BBox-Label-Tool/main.py", line 142, in loadDir
self.category = int(s)
ValueError: invalid literal for int() with base 10: ''
////////////////////////////////////////////////////
Doesn't it work on windows?
@puzzledqs ,Could you help me solve this?
I had to edit main.py lines 10-11 to:
from tkinter import *
import tkinter.messagebox
to work with newer versions of Python.
Is it possible to let some images unlabeled ? For example, if you have certain images where there is no object to label, is it possible to have the format :
1
0 0 0 0
I am not sure about how to write that being not familiar with "event". I'd like to have a default "(0 0) -> (0 0)" meaning in the text file (line 1 : 1 ; line 2 : 0 0 0 0), so that I can just click on "Next" and save the picture labeled like that because I have a lot of images without interesting objects. However, when there are objects, I'd like to cancel this "default" box, by just clicking somewhere in the picture and save the coordinates of the boxes in the text file.
Any leads on how to do that ?
Thank you for the amazing tool.
Is it possible to give markups on edges of the drawn bounding box and resize it after. I want to do this instead of removing and redrawing a new bounding box.
When I draw a bounding box in main.py, the ROI the boudning box applied is a little small.
I just wanna increase the loaded image size for more accurate bounding box. is there a tool or method in main.py program? (or do I write some code for this?
I have images with those size 1944, 2592. which part of the code should I modify for displaying the whole image?
thanks in advance.
olive
and teal
are not accepted color names. The third and fourth bounding boxes don't show up as a result of this. main.py line 19 need to be modified to include valid Tkinter colors.
Please add license, so that the terms of usage of this tool are clear.
test@virtualboxmachine:~/Downloads/BBox-Label-Tool-master$ python main.py
While executing this command, the following error occurred as follows:
Exception in Tkinter callback
Traceback (most recent call last):
File "/usr/lib/python2.7/lib-tk/Tkinter.py", line 1540, in call
return self.func(*args)
File "main.py", line 126, in loadDir
self.category = int(s)
ValueError: invalid literal for int() with base 10: ''
Kindly help me to resolve this issue.
i can't load my picture
s = r'D:\workspace\python\labelGUI' <= is that image's path??
coz I use in anaconda and i wander know how to set the image's path.
Hello,
I just used jxgu1016/BBox-Label-Tool-Multi-Class to classify multiple different chickens in each of 200 images. The images have up to 6 different chickens each. I would like to now use format.py to prepare the labels for a YOLO model. The problem is that format.py does not expect the label format that jxgu1016/BBox-Label-Tool-Multi-Class creates. How can I use this new multi-line label format to train a new YOLO object detection model? Thank you.
This is the format of the jxgu1016/BBox-Label-Tool-Multi-Class labels for a single image:
4
66 201 236 452 Kielyr
421 235 578 369 Vespyr
653 217 885 376 Bo
918 159 1156 617 Zaja
format.py is here: https://github.com/chuanenlin/autoturk/blob/master/format.py
Hi, I am on Macbook pro and found it keep blinking while moving mouse cursor over the panel, why?
I'm trying to load images from a directory into the LabelTool, but everytime I try I get "No .jpeg images found in the specified dir!". Even running with the test images provided in the download does not work. The script seems to be looking in the right location:
Code modified to show above screenshot is at https://pastebin.com/Y1BFWww2
My 001 folder is as follows:
I am using a fork to work with high res images from @toobaimt found at #25
Any help would be appreciated.
Hello,
Why I can't annotate more than 2 objects in one image?
The third box doesn't want to show herself and aren't reported in .txt file.
I'm trying to load some images, and any time they load, they are extremely blown up and cut off. The images I'm trying to load are all 256x256 pixels. Here's an example of what it looks like in the app:
The image I'm trying to work with is a full shot of a pill/prescription bottle, and it only shows one corner of the image, blown up. Any idea on how to fix this?
Hello,
I just used jxgu1016/BBox-Label-Tool-Multi-Class to classify multiple different chickens in each of 200 images. The images have up to 6 different chickens each. I would like to now use format.py to prepare the labels for a YOLO model. The problem is that format.py does not expect the label format that jxgu1016/BBox-Label-Tool-Multi-Class creates. How can I use this new multi-line label format to train a new YOLO object detection model? Thank you.
This is the format of the jxgu1016/BBox-Label-Tool-Multi-Class labels for a single image:
4
66 201 236 452 Kielyr
421 235 578 369 Vespyr
653 217 885 376 Bo
918 159 1156 617 Zaja
format.py is here: https://github.com/chuanenlin/autoturk/blob/master/format.py
where can i find these modules...
As you can see below, I succeeded to label in imdir:001
But, when I loaded imdir:002, I couldn't bring jpg files into GUI program('main.py')
Also, I uploaded main.txt file. (I think that this file is same as main.py except "path"
main.txt
when i try to press the load button, in command line shows error like this
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Python27\lib\lib-tk\Tkinter.py", line 1541, in __call__
return self.func(*args)
File "main.py", line 126, in loadDir
self.category = int(s)
ValueError: invalid literal for int() with base 10: ''
anyone can help me solve this problem ?
I have an application of labeling number plates of bikes in which the images are skewed, but we are able to make only rectangular bounding boxes as is allowed to be taken by YOLO . So the area of interest also includes some unwanted regions. Any suggestions on handling such cases?
Examples of such cases are attached
I've only ever used BBox before on Linux, and, now that I've tried to install it on a Windows OS, I've been experience a lot of difficulties in just opening the GUI. As soon as I input
open(exec("main.py").read())
into the Python IDLE, nothing happens. No error messages or anything. Keep in mind that I have changed the current working directory at this point. Is it possible that I don't have necessary libraries installed? Can the software adequately run on Windows?
Thanks for any help; I really appreciate it.
When I start the program and press load, python returns the following error
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\user\Anaconda3\envs\yolo_training\lib\lib-tk\Tkinter.py", line 1541, in call
return self.func(*args)
File "main.py", line 127, in loadDir
self.category = int(s)
ValueError: invalid literal for int() with base 10: ''
I am running the program on conda environment running on python2.7
Traceback (most recent call last):
File "/usr/lib/python2.7/lib-tk/Tkinter.py", line 1540, in call
return self.func(*args)
File "main.py", line 126, in loadDir
self.category = int(s)
ValueError: invalid literal for int() with base 10: ''
Large resolution photos do not fit the screen at multi-class branch
window doesn't fit in Ubuntu screen while we are creating boxes
is there a way to change the format that can be accepted by yolov2? i am asking for multi class image. Thank you
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.