Labelme labeled image segmentation (semantic segmentation) samples

1. Download: pip install labelme

2. Save the image to be marked into the images directory, and cd to the parent directory of the images directory

3. Create a new labels.txt file, ansi code, and enter the category

__ignore__
_background_
aeroplane
bicycle
bird
boat
bottle
bus
car
cat
chair
cow
diningtable
dog
horse
motorbike
person
potted plant
sheep
sofa
train
tv/monitor

4. Start labelme from the command line

labelme images --labels labels.txt --nodata

After marking all pictures, exit labelme

5. Download the batch conversion py code

enterhttps://github.com/wkentaro/labelme/tree/master/examples/semantic_segmentationPage, downloadlabelme2voc.py

6. Conversion

python.exe ./labelme2voc.py images data_dataset_voc --labels ./labels.txt

The effect after conversion is as follows:

Intelligent Recommendation

Labelme Semantic Segmentation Tooling Installation and Use

Environment: ubuntu18.04 (with python2.7 and python3.6) +conda (3.7 downgrade to 3.6, that is, install python3.6) I took the ubuntu installation method: Configure pip3: sudo apt install pyhton3-pip In...

python labelme semantic segmentation of production data collection

Specific process is as follows 1 labelme divided semantic annotation data to generate label file .json 2 labelme_json_to_dataset .json file (labelme label file generated after)...

Semantic segmentation learning series (1) Installation of labelme

Foreword Labelme is the software that labels the sample data before the semantic/instance segmentation model training, just like labelimage is the labeling software for target detection training data....

Semantic segmentation data set production labelme

1. First install labelme: open the Anaconda prompt-->pip install labelme-->Enter (If an error is reported during the installation, please don't give up just a few more attempts! Because you may ...

Minconda installs labelme semantic segmentation annotation

The first step is to download minconda, here I use the software site of Tsinghua University, it is really fasthttps://mirror.tuna.tsinghua.edu.cn/ The second installment of minconda is to just click t...