ImageLabeler

基于Qt实现的图片数据标注工具. Image Annotation Tool Based on Qt, supporting 2D/3D detection/segmentation.

Jialong Wu

Introduction

This project uses Qt to develop an image data annotation tool with a graphical interface for data preparation in the field of artificial intelligence, specifically for users to annotate 2D and 3D images in two ways:

1. Detection annotation: mark the smallest possible rectangular (cube) box covering an object in the image and add a label to the box.
2. Segmentation annotation: indicates the pixels (voxels) covered by the object in the image, and also adds a label to the shape.

For details, please refer to our report (in Chinese).

Environment

Qt 5.12.4, Qt Creator 4.9.2

Installation

Download link for Windows Application: Google Drive.

Usage

We refer users to Section 3 in our project report (in Chinese) for detailed instructions for use.

Snapshots

Acknowledgement

We are very grateful for the reference, labelme.

This is the course project for Practical Training For Programming in School of Software, Tsinghua University, instructed by Prof. Xu. We would like to thank Prof. Xu and TAs for their devotion to the course.

References

1. labelme
2. Qt document
3. Random colors Qt
4. QDarkStyleSheet