Skip to content

Real time automatic creation of bounding boxes as well as salience maps around detected objects for fixed backgrounds. Save images as well as the corresponding annotations once the object has been detected. Uses OpenCV and TensorFlow

License

Notifications You must be signed in to change notification settings

shreyas-bk/BOXER_auto_annotation_tool

Repository files navigation

BOXER_auto_annotation_tool

Real time automatic creation of bounding boxes as well as salience maps around detected objects for fixed backgrounds.

No model training required as it only uses difference in feature maps between the image of the object and the image of the background, hence it can run in real-time.

Google Colab Link for tutorial

Click herw to open the notebook

Examples

Bounding box creation around detected object:

example

Bounding box creation around detected object with salience map highlighted:

example

Requirements:

  • cv2
  • tensorflow

Notes:

  • Salience maps are optional (can be toggled)
  • Images with bounding box around the object as well as a text file for the annotation can be saved
  • Salience maps works best with light colored, stable backgrounds

About

Real time automatic creation of bounding boxes as well as salience maps around detected objects for fixed backgrounds. Save images as well as the corresponding annotations once the object has been detected. Uses OpenCV and TensorFlow

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published