Skip to content
This repository has been archived by the owner on Aug 29, 2024. It is now read-only.
/ pnno Public archive

dataset or annotation file format conversion

License

Notifications You must be signed in to change notification settings

zjykzj/pnno

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Language: 🇺🇸 🇨🇳

This repository is no longer maintained. For updated content, please visit: zjykzj/vocdev and zjykzj/cocodev

«pnno» realizes the conversion of dataset or annotation data in different formats

This repo provides several script tools for the following aspects:

  1. convert different data sets into the format specified by the algorithm. Currently implemented
    1. LabelImg annotation -> YoloV5 data format
    2. VisDrone data set -> TLT KITTI data format
  2. convert PyTorch ImageFolder dataset to LMDB format
  3. show image in command line

Table of Contents

Background

The processing of data sets is often involved in the process of algorithm development, which needs to be converted into the format specified in algorithm training. Many scripts are accumulated over time. Whether these programs can be integrated together can not only be reused, but also reduce the difficulty of the next implementation. Let's Do It ! ! !

Install

$ pip install pnno

Usage

Dataset/Label Convert

Basic operation as follows:

$ pnno -f <cfg_file>

Operation 1: convert tzutalin/labelImg label file to ultralytics/yolov5 specified dataset format. Refer to the configuration file configs/labelimg_2_yolov5.yaml

Operation 1: convert VisDrone/VisDrone-Dataset dataset to KITTI label format. Refer to the configuration file configs/visdrone_2_tlt.yaml

Operation 3: convert PyTorch ImageFolder dataset to LMDB format. Refer to demo/imagenet_lmdb

For more usage, refert to demo/

Show Image in Command Line

$ simg -f <IMG_FILE>

Maintainers

  • zhujian - Initial work - zjykzj

Contributing

Anyone's participation is welcome! Open an issue or submit PRs.

Small note:

License

Apache License 2.0 © 2020 zjykzj