convert pascal voc dataset to yolo format

convert_voc_to_yolo.md

Convert PascalVOC Annotations to YOLO

This script reads PascalVOC xml files, and converts them to YOLO txt files.

Note: This script was written and tested on Ubuntu. YMMV on other OS's.

Disclaimer: This code is a modified version of Joseph Redmon's voc_label.py

Instructions:

1. Place the convert_voc_to_yolo.py file into your data folder.
2. Edit the dirs array (line 8) to contain the folders where your images and xmls are located. Note: this script assumes all of your images are .jpg's (line 13).
3. Edit the classes array (line 9) to contain all of your classes.
4. Run the script. Upon running the script, each of the given directories will contain a 'yolo' folder that contains all of the YOLO txt files. A text file containing all of the image paths will be created in the cwd, for each given directory.

Make sure to put images and xml files in the root of train.Like this(image is in comment),here my folder name is VOCData and yolo folder is generated by script.

convert_voc_to_yolo.py:

import glob
import os
import pickle
import xml.etree.ElementTree as ET
from os import listdir, getcwd
from os.path import join

dirs = ['train', 'val']
classes = ['person', 'car']

def getImagesInDir(dir_path):
    image_list = []
    for filename in glob.glob(dir_path + '/*.jpg'):
        image_list.append(filename)

    return image_list

def convert(size, box):
    dw = 1./(size[0])
    dh = 1./(size[1])
    x = (box[0] + box[1])/2.0 - 1
    y = (box[2] + box[3])/2.0 - 1
    w = box[1] - box[0]
    h = box[3] - box[2]
    x = x*dw
    w = w*dw
    y = y*dh
    h = h*dh
    return (x,y,w,h)

def convert_annotation(dir_path, output_path, image_path):
    basename = os.path.basename(image_path)
    basename_no_ext = os.path.splitext(basename)[0]

    in_file = open(dir_path + '/' + basename_no_ext + '.xml')
    out_file = open(output_path + basename_no_ext + '.txt', 'w')
    tree = ET.parse(in_file)
    root = tree.getroot()
    size = root.find('size')
    w = int(size.find('width').text)
    h = int(size.find('height').text)

    for obj in root.iter('object'):
        difficult = obj.find('difficult').text
        cls = obj.find('name').text
        if cls not in classes or int(difficult)==1:
            continue
        cls_id = classes.index(cls)
        xmlbox = obj.find('bndbox')
        b = (float(xmlbox.find('xmin').text), float(xmlbox.find('xmax').text), float(xmlbox.find('ymin').text), float(xmlbox.find('ymax').text))
        bb = convert((w,h), b)
        out_file.write(str(cls_id) + " " + " ".join([str(a) for a in bb]) + '\n')

cwd = getcwd()

for dir_path in dirs:
    full_dir_path = cwd + '/' + dir_path
    output_path = full_dir_path +'/yolo/'

    if not os.path.exists(output_path):
        os.makedirs(output_path)

    image_paths = getImagesInDir(full_dir_path)
    list_file = open(full_dir_path + '.txt', 'w')

    for image_path in image_paths:
        list_file.write(image_path + '\n')
        convert_annotation(full_dir_path, output_path, image_path)
    list_file.close()

    print("Finished processing: " + dir_path)

Please help. I have uploaded my entire dataset to google drive and mounted it to Google Colab. I am trying to run this script on Colab. My dataset directory is as follows: There are two separate folders for Images and Annotations. Both of these folders reside in a folder named MyDataset. I am compiling this script in MyDataset folder but I am getting an empty yolo folder and .txt file. Also, I don't have train and validation folders of my dataset yet. Any help would be highly appreciated!

work well. thanks

@c-pineau Thanks for your post. I had the same issue but I am not getting the output yolo file. I am wonder about the directory structure. I am working with colab. Also my data have not train and val split just like @mhamza19 commented above. What to do in this situation? Your guidance will be appreciated.

@imadgohar What do you mean by Yolo file? Also, I would suggest you to use Roboflow as It will save you a lot of time while automating all this data-related stuff in a few clicks.

@imadgohar What do you mean by Yolo file? Also, I would suggest you to use Roboflow as It will save you a lot of time while automating all this data-related stuff in a few clicks.

I am done with it thanks.

import glob
import os
import pickle
import xml.etree.ElementTree as ET
from os import listdir, getcwd
from os.path import join

dirs = ['train', 'test']
classes = ['apple', 'banana', 'orange']


def getImagesInDir(dir_path):
    image_list = []
    for filename in glob.glob(dir_path + '\\*.jpg'):
        image_list.append(filename)

    return image_list

def convert(size, box):
    if size[0] == 0:
        dw = 1./(size[0]+0.00001)
    else:
        dw = 1./(size[0])
        
    if size[0] == 0:
        dh = 1./(size[1]+0.00001)
    else:
        dh = 1./(size[1])

    x = (box[0] + box[1])/2.0 - 1
    y = (box[2] + box[3])/2.0 - 1
    w = box[1] - box[0]
    h = box[3] - box[2]
    x = x*dw
    w = w*dw
    y = y*dh
    h = h*dh
    return (x,y,w,h)

def convert_annotation(dir_path, output_path, image_path):
    basename = os.path.basename(image_path)
    basename_no_ext = os.path.splitext(basename)[0]

    in_file = open(dir_path + '\\' + basename_no_ext + '.xml')
    out_file = open(output_path + basename_no_ext + '.txt', 'w')
    tree = ET.parse(in_file)
    root = tree.getroot()
    size = root.find('size')
    w = int(size.find('width').text)
    h = int(size.find('height').text)

    for obj in root.iter('object'):
        difficult = obj.find('difficult').text
        cls = obj.find('name').text
        if cls not in classes or int(difficult)==1:
            continue
        cls_id = classes.index(cls)
        xmlbox = obj.find('bndbox')
        b = (float(xmlbox.find('xmin').text), float(xmlbox.find('xmax').text), float(xmlbox.find('ymin').text), float(xmlbox.find('ymax').text))
        bb = convert((w,h), b)
        out_file.write(str(cls_id) + " " + " ".join([str(a) for a in bb]) + '\n')

cwd = getcwd()
print('not working')
for dir_path in dirs:
    full_dir_path = cwd + '\\' + dir_path
    output_path = full_dir_path +'\\yolo\\'

    if not os.path.exists(output_path):
        print(output_path)
        os.makedirs(output_path)

    image_paths = getImagesInDir(full_dir_path)
    list_file = open(full_dir_path + '.txt', 'w')

    for image_path in image_paths:
        print(image_path)
        list_file.write(image_path + '\n')
        convert_annotation(full_dir_path, output_path, image_path)
    list_file.close()

    print("Finished processing: " + dir_path)
print('gihub.com/ipiyushvaghela')

For Windows users...

Suppose the DATASET folder contains the train and test folders...
then just create a .py file inside the DATASET folder and paste the above code. and run that py file your .txt will be created inside train --> yolo.

import glob
import os
import pickle
import xml.etree.ElementTree as ET
from os import listdir, getcwd
from os.path import join

dirs = ['train', 'test']
classes = ['apple', 'banana', 'orange']


def getImagesInDir(dir_path):
    image_list = []
    for filename in glob.glob(dir_path + '\\*.jpg'):
        image_list.append(filename)

    return image_list

def convert(size, box):
    if size[0] == 0:
        dw = 1./(size[0]+0.00001)
    else:
        dw = 1./(size[0])
        
    if size[0] == 0:
        dh = 1./(size[1]+0.00001)
    else:
        dh = 1./(size[1])

    x = (box[0] + box[1])/2.0 - 1
    y = (box[2] + box[3])/2.0 - 1
    w = box[1] - box[0]
    h = box[3] - box[2]
    x = x*dw
    w = w*dw
    y = y*dh
    h = h*dh
    return (x,y,w,h)

def convert_annotation(dir_path, output_path, image_path):
    basename = os.path.basename(image_path)
    basename_no_ext = os.path.splitext(basename)[0]

    in_file = open(dir_path + '\\' + basename_no_ext + '.xml')
    out_file = open(output_path + basename_no_ext + '.txt', 'w')
    tree = ET.parse(in_file)
    root = tree.getroot()
    size = root.find('size')
    w = int(size.find('width').text)
    h = int(size.find('height').text)

    for obj in root.iter('object'):
        difficult = obj.find('difficult').text
        cls = obj.find('name').text
        if cls not in classes or int(difficult)==1:
            continue
        cls_id = classes.index(cls)
        xmlbox = obj.find('bndbox')
        b = (float(xmlbox.find('xmin').text), float(xmlbox.find('xmax').text), float(xmlbox.find('ymin').text), float(xmlbox.find('ymax').text))
        bb = convert((w,h), b)
        out_file.write(str(cls_id) + " " + " ".join([str(a) for a in bb]) + '\n')

cwd = getcwd()
print('not working')
for dir_path in dirs:
    full_dir_path = cwd + '\\' + dir_path
    output_path = full_dir_path +'\\yolo\\'

    if not os.path.exists(output_path):
        print(output_path)
        os.makedirs(output_path)

    image_paths = getImagesInDir(full_dir_path)
    list_file = open(full_dir_path + '.txt', 'w')

    for image_path in image_paths:
        print(image_path)
        list_file.write(image_path + '\n')
        convert_annotation(full_dir_path, output_path, image_path)
    list_file.close()

    print("Finished processing: " + dir_path)
print('gihub.com/ipiyushvaghela')

For Windows users...

Suppose the DATASET folder contains the train and test folders... then just create a .py file inside the DATASET folder and paste the above code. and run that py file your .txt will be created inside train --> yolo.

it didn't work for me

very nice
It worked for me
tnks bro

Very nice post!!! It worked well for me.

very nice, thank you for share!

it didn't work for me. i got this error.
runfile('D:/bit dataset file/BITVehicle_Dataset/from_xml_to_yolo_convert_label_script.py', wdir='D:/bit dataset file/BITVehicle_Dataset')
Traceback (most recent call last):

File ~.conda\envs\thesis_two\lib\site-packages\spyder_kernels\py3compat.py:356 in compat_exec
exec(code, globals, locals)

File d:\bit dataset file\bitvehicle_dataset\from_xml_to_yolo_convert_label_script.py:75
convert_annotation(full_dir_path, output_path, image_path)

File d:\bit dataset file\bitvehicle_dataset\from_xml_to_yolo_convert_label_script.py:47 in convert_annotation
w = int(size.find('width').text)

ValueError: invalid literal for int() with base 10: '[[1920]]'

Because my labeling files consist like that size.

[[1920]]
[[1080]]
3
How can i change this code? Thank you.

it didn't work for me. i got this error. runfile('D:/bit dataset file/BITVehicle_Dataset/from_xml_to_yolo_convert_label_script.py', wdir='D:/bit dataset file/BITVehicle_Dataset') Traceback (most recent call last):

File ~.conda\envs\thesis_two\lib\site-packages\spyder_kernels\py3compat.py:356 in compat_exec exec(code, globals, locals)

File d:\bit dataset file\bitvehicle_dataset\from_xml_to_yolo_convert_label_script.py:75 convert_annotation(full_dir_path, output_path, image_path)

File d:\bit dataset file\bitvehicle_dataset\from_xml_to_yolo_convert_label_script.py:47 in convert_annotation w = int(size.find('width').text)

ValueError: invalid literal for int() with base 10: '[[1920]]'

Because my labeling files consist like that size. [[1920]] [[1080]] 3 How can i change this code? Thank you.

This code worked for me but I used another code for more convenience.
https://github.com/Ryo-Kawanami/xml2yolo/tree/master

The author's code works. In VOC dataset, tag xml files and jpg images are stored separately. You need to place your xml tag file and jpg image in the same folder and then run the program to get the result of the conversion. Ensuring file paths as:
├──convert_voc_to_yolo.py
└──train
└── xxx0.xml
└── xxx0.jpg
└──test
└── xxx1.xml
└── xxx1.jpg