isabek / xmltotxt Goto Github PK

View Code? Open in Web Editor NEW

141.0 2.0 40.0 16 KB

ImageNet file xml format to Darknet text format

Python 100.00%

darknet-image-classification imagenet-dataset imagenet-classifier

xmltotxt's Introduction

XmlToTxt

ImageNet file xml format to Darknet text format. Works well with directories and subdirectories.

Installation

sudo pip install -r requirements.txt

Usage

python xmltotxt.py -c cls.txt -xml xml -out out

Mandatory arguments

-xml

Optional arguments

-c, -out

Example

Input xml file.

<annotation>
	<filename>image-0000016.jpg</filename>
	<size>
		<width>1920</width>
		<height>1080</height>
	</size>
	<object>
		<name>sedan</name>
		<bndbox>
			<xmin>75</xmin>
			<ymin>190</ymin>
			<xmax>125</xmax>
			<ymax>210</ymax>
		</bndbox>
	</object>
</annotation>

Output text file.

5 0.052083 0.185185 0.026042 0.018519

Motivation

I used Darknet for real-time object detection and classification. Sometimes you need to collect your own training dataset for train your model. I collected training dataset images and fine awesome tool for labeling images. But it generates xml files. So I needed to implement tool which translates from ImageNet xml format to Darknet text format. Also compatible with latest YOLOv5 by Ultralytics.

xmltotxt's People

Contributors

Stargazers

Watchers

xmltotxt's Issues

Citation for this repo

Hi there,

I'm looking for a citation of this repo with bitex format for my thesis. Thanks so much!!

ERROR:roor:('Missing required root aggregate "annotation" ',)

Support custom classes file

Need to set classes file as an optional argument.

python xmltotxt.py -c custom_classes.txt -xml xml -out out

Conversion Error : ZeroDivisionError: float division by zero

ERROR:root:('Missing required array: "objects"',)
Traceback (most recent call last):
File "xmltotxt.py", line 33, in
main()
File "xmltotxt.py", line 29, in main
transformer.transform()
File "/home/MyPc/Downloads/XmlToTxt/transformer.py", line 18, in transform
self.write_to_txt(annotations, classes)
File "/home/MyPc/Downloads/XmlToTxt/transformer.py", line 23, in write_to_txt
f.write(self.to_darknet_format(annotation, classes))
File "/home/MyPc/Downloads/XmlToTxt/transformer.py", line 31, in to_darknet_format
x, y, width, height = self.get_object_params(obj, annotation.size)
File "/home/MyPc/Downloads/XmlToTxt/transformer.py", line 48, in get_object_params
x = absolute_x / image_width
ZeroDivisionError: float division by zero

Can You Help me About this issue ? Thanks

ERROR:root:('Invalid numeric value "453.0395" at annotation/object[0]/bndbox/xmin',)

I GOT above error when tried to convert Hollywoodheads dataset xml file into yolo txt format. error occurs at both xmin and xmax value

output broken

running the base example only creates 1 file which also happens to be empty:

/dataSets/XmlToTxt$ python xmltotxt.py -xml xml -out out
Please, add 'sedan' to classes.txt file.
/dataSets/XmlToTxt$ ls out
image-0000001.txt

KeyError: 'person'

Traceback (most recent call last):
File "xmltotxt.py", line 33, in
main()
File "xmltotxt.py", line 29, in main
transformer.transform()
File "/home/ubuntu/ext/XmlToTxt/transformer.py", line 18, in transform
self.write_to_txt(annotations, classes)
File "/home/ubuntu/ext/XmlToTxt/transformer.py", line 23, in write_to_txt
f.write(self.to_darknet_format(annotation, classes))
File "/home/ubuntu/ext/XmlToTxt/transformer.py", line 29, in to_darknet_format
result.append("%d %.6f %.6f %.6f %.6f" % (classes[obj.name], x, y, width, height))
KeyError: 'person'

Didn't work for Korean characters (UnicodeDecodeError: 'charmap' codec can't decode byte 0x81 in position 7: character maps to <undefined>)

Hi Isabek,

I'm really thankful for you making this repo. However it didn't work for Korean characters, or I couldn't make it? Could you please check and update it (if possible). Here is the error when I'm running the code:

Traceback (most recent call last):
  File "xmltotxt.py", line 33, in <module>
    main()
  File "xmltotxt.py", line 29, in main
    transformer.transform()
  File "D:\Projects\XmlToTxt\transformer.py", line 15, in transform
    classes = reader.get_classes()
  File "D:\Projects\XmlToTxt\reader.py", line 18, in get_classes
    lines = f.readlines()
  File "C:\Users\Scott\Anaconda3\envs\xmltotxt\lib\encodings\cp1252.py", line 23, in decode
    return codecs.charmap_decode(input,self.errors,decoding_table)[0]
UnicodeDecodeError: 'charmap' codec can't decode byte 0x81 in position 7: character maps to <undefined>

I'm guessing it might be an encoding issue. Thanks in advance!

ERROR:root:('Missing required array: "objects"',)

HI I got this error while running this code

ERROR:root:('Missing required array: "objects"',)

my .XMl file is like this

<?xml version="1.0"?>

-<annotation>

<folder>None</folder>

<filename>train/images/fruit_0_1.png</filename>

<path>None</path>


-<source>

<database>Unknown</database>

</source>


-<size>

<width>250</width>

<height>250</height>

<depth>3</depth>

</size>

<segmented>0</segmented>


-<object>

<name>oranges</name>

<pose/>

<truncated>0</truncated>

<difficult>0</difficult>


-<bndbox>

<xmin>56</xmin>

<ymin>106</ymin>

<xmax>73</xmax>

<ymax>123</ymax>

</bndbox>

</object>

</annotation>

Sedan example and sample output wrong

The sedan is class of index 5 but in all samples it is indexed as 4