[Object detection] Install image annotation tool labelme under Windows

Detailed installation process reference: https://github.com/wkentaro/labelme

1 Installation of Labelme under Windows

Execute the following commands:

1) 打开Anaconda命令行工具
2) 执行 conda create --name=labelme python=3.6
3) 执行activate labelme
4) 执行 conda install pyyaml
5) 执行 pip install labelme
6) 执行 labelme

2 Use of Labelme

1) Click "open" to open the image to be annotated, and select to annotate the target area. Click after the annotation is completed to generate a json file.
2) Enter the labelme installation directory and execute python labelme_json_to_dataset <文件名>.json. You can get a folder with five files, namely: *.png, info.yaml , label.png, label_viz.png，label_names.txt. The label.pngsum info.yamlis what we need to use, which label.pngis equivalent to the mask file.

The generated files are as follows:

---

Appendix I

I modified the labelme_json_to_datasetfile a bit, and it can be generated in batches, python <文件夹名称>just :

labelme_json_to_dataset is modified as follows:

#!D:\conda\envs\labelme\python.exe

import argparse
import json
import os
import os.path as osp
import warnings

import numpy as np
import PIL.Image
import yaml

from labelme import utils


def main():
    parser = argparse.ArgumentParser()
    parser.add_argument('json_file')
    parser.add_argument('-o', '--out', default=None)
    args = parser.parse_args()

    json_file = args.json_file

    list = os.listdir(json_file) 
    for i in range(0, len(list)):
        path = os.path.join(json_file, list[i])
        if os.path.isfile(path):
            data = json.load(open(path))
            img = utils.img_b64_to_array(data['imageData'])
            lbl, lbl_names = utils.labelme_shapes_to_label(img.shape, data['shapes'])

            captions = ['%d: %s' % (l, name) for l, name in enumerate(lbl_names)]
            lbl_viz = utils.draw_label(lbl, img, captions)
            out_dir = osp.basename(list[i]).replace('.', '_')
            out_dir = osp.join(osp.dirname(list[i]), out_dir)
            if not osp.exists(out_dir):
                os.mkdir(out_dir)

            PIL.Image.fromarray(img).save(osp.join(out_dir, 'img.png'))
            PIL.Image.fromarray(lbl).save(osp.join(out_dir, 'label.png'))
            PIL.Image.fromarray(lbl_viz).save(osp.join(out_dir, 'label_viz.png'))

            with open(osp.join(out_dir, 'label_names.txt'), 'w') as f:
                for lbl_name in lbl_names:
                    f.write(lbl_name + '\n')

            warnings.warn('info.yaml is being replaced by label_names.txt')
            info = dict(label_names=lbl_names)
            with open(osp.join(out_dir, 'info.yaml'), 'w') as f:
                yaml.safe_dump(info, f, default_flow_style=False)

            print('Saved to: %s' % out_dir)
if __name__ == '__main__':
    main()

Appendix II

labelmeThe generated mask tag label.pngis stored in 16 bits, and opencv8 bits are read by default, and 16 bits need to be converted to 8 bits. You can use the following Java code to convert:

Someone has provided a method for C++ conversion: http://blog.csdn.net/l297969586/article/details/79154150

    /**
     * 说明 : 16位 转化为 8位
     */
    public void transform(String originPath, String transformPath) {

        if(!new File(originPath).exists()) {
            System.out.println("文件不存在！");
            return;
        }

        //原图片
        Mat src = Imgcodecs.imread(originPath, Imgcodecs.CV_LOAD_IMAGE_UNCHANGED);

        //生成的图片
        Mat mat = Mat.zeros(src.rows(), src.cols(), CvType.CV_8SC1);
        for (int r = 0; r < src.rows(); r++) {
            for(int c = 0; c < src.cols(); c++){
                mat.put(r, c, src.get(r, c));
            }
        }

        Imgcodecs.imwrite(transformPath, mat);

    }