当前位置:网站首页>Labelme for image data annotation

Labelme for image data annotation

2022-06-11 18:23:00 Yalin melon seeds

scene

Use Labelme The tool labels the picture data .

step

install Labelme

Go to this address https://github.com/wkentaro/labelme/releases Download the corresponding installation package .
After downloading , After installation .
Labelme Of app

label.txt

Construct a Labelme Know the classification file

__ignore__
_background_
cat

Use Labelme mark

hypothesis cats2 In file 10 This picture is a file that needs to be marked , Here's the picture :
 Files to be annotated
Use the following command , start-up Labelme Annotate .

/Applications/Labelme.app/Contents/MacOS/labelme --labels labels.txt --nodata

choice Open Dir open cat2 Folder , And then click Create Polygons To start labeling the image data .
Open Dir Folder
Create Polygons
Last , The next step is to label the picture , Use label2voc.py Script for data conversion .

python3 label2voc.py cats2 cats_output --label label.txt 

label2voc.py source :
https://github.com/wkentaro/labelme/blob/main/examples/semantic_segmentation/labelme2voc.py

The generated folder is shown in the following figure :
label2voc result

原网站

版权声明
本文为[Yalin melon seeds]所创,转载请带上原文链接,感谢
https://yzsam.com/2022/162/202206111804320948.html