Skip to content

henniekim/pascal_voc_data_parsing

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 

Repository files navigation

PASCAL VOC DATA PARSING

[한국어]
 PASCAL VOC 데이터 셋은 AnnotaionsJPEGimage로 나누어져 있습니다.
AnnotationXML 명세로 되어있으며, 각 물체의 클래스와 바운딩 박스 좌표로 이루어져 있습니다.
분류기 모델을 학습시키기 위해서는 각 바운딩 박스 안에 있는 사진과 레이블을 가져와야 합니다.

본 프로젝트는 PASCAL VOC 데이터를 BeautifulSoup 과 opencv 라이브러리를 이용하여 가공합니다. 여러 이미지가 들어있는 데이터 셋에서 각 물체를 분리하여 224 x 224 크기로 변환을 하여 저장합니다. 레이블은 [20 x 1] one hot encoding 으로 텍스트 파일로 저장합니다.

 여기서 만든 데이터를 이용하여, 학습을 시키고자 하는 분은 https://github.com/henniekim/python_keras_vgg_16 을 방문해주세요.

감사합니다.

[ENGLISH]
PASCAL VOC data set consist of "Annotations" and "JPEGimage". Annotations are written in XML format which contain each object class and bounding box coordinates. You have to bring each bounding box and its photo and label in order to train the network.

This project processes on PASCAL VOC data with BeautifulSoup and opencv libs. Each photo contains many objects. It detaches each object, resizes to 224 x 224 and save to local storage. Label is saved in *.txt with [ 20 x 1 ] one hot encoding.

Anyone who want to train with the dataset made hers, visit https://github.com/henniekim/python_keras_vgg_16

Thank you !

설치 방법

적당한 폴더에 PASCAL VOC 데이터를 받습니다.
데이터는 다음 미러 사이트에서 내려받을 수 있습니다.

https://pjreddie.com/projects/pascal-voc-dataset-mirror/

Download PASCAL VOC dataset to proper folder. You can get it from as below.

https://pjreddie.com/projects/pascal-voc-dataset-mirror/

사용 예제

터미널 창에서 다음과 같이 실행하세요.

python3 voc_dataset_parsing_edited.py --dataPath $YOUR_DATA_PATH --years 2007 2008 2009 

개발 환경 설정

Python 3.x 버젼과, BeautifulSoup4, Opencv가 필요합니다.

pip3 install BeautifulSoup4
pip3 install opencv-python

업데이트 내역

  • 0.0.1
    • 작업 진행 중
  • 0.0.2
    • Add english document
  • 0.0.3
    • 이제 특정 연도를 선택하여 데이터를 처리할 수 있습니다.

추후 업데이트 예정

  • 처리한 데이터의 파일명에 연도를 포함하는 옵션 추가하기

정보

김동현 – seru_s@me.com

라이센스

MIT © henniekim

Releases

No releases published

Packages

 
 
 

Languages