A customized version of VIA image annotator which supports server deployment (python-flask backend).
在VIA的图像标注工具上做了升级,增加了flask后端。没用数据库,只配置文件夹来管理数据。增加了自动保存和逐张加载的功能。
A lightweight version for deployment image annotator on server. Without database plugin, it only uses a config file with directory paths. Auto-save and download images one by one when browse.
flask, tqdm, yaml, PIL, etc.
learn from VIA image annotator version 3 https://www.robots.ox.ac.uk/~vgg/software/via/
run with script run.sh
2021.08.31, 1. 增加了数据的users属性,根据用户名显示可用的数据
2. 修复了select选择图片未能自动加载的bug
3. 增加了voc与本数据格式(via)之间的转换工具
2021.09.08, 1. 增加了旋转图片的功能(服务器数据同步改变)
- 在服务器端的每个样本标注单独保存 (Save annotation file separately)
- 支持一键转化输出coco等格式 (Support convert annotation file to coco or any other formats by one-click)
- 支持图片分布在多个目录(Support read images of one dataset from multi-directory)