Skip to content

Latest commit

 

History

History
52 lines (29 loc) · 1.81 KB

README.md

File metadata and controls

52 lines (29 loc) · 1.81 KB
Raw

Face Detection & Contour Extraction

My realization of face detection & contouring based on Haar-Cascade classifier and combination of simple digital image processing methods.

Content in Each Folder

code

  • Folder opencv_xml includes all the pre-trained xml files provided by cv2 library.
  • main.py - aggregated file for the realization of the whole algorithm, contains test_on_camera function.
  • evaluation.py - makes evaluation of the algorithm, and plots of the results.
  • my_canny_detector.py - implementation of canny method of contouring.
  • my_fg_bg_detector.py - implementation of foreground / background method of contouring.
  • my_haar_cascade.py - implementation of Haar-Cascade classifier.

images

  • my_dataset

    Includes 10 image selected from FEI database.

  • test_imgs

    Includes test images for the self-use and plots.

  • haar_results

    Includes Haar-Cascade classifier results on my_dataset.

  • contouring_results

    Includes contouring results from my algorithm on haar_results.

  • contouring_compare

    Includes manually drew contouring on haar_results

  • figures

    Includes evaluation figures for the report.

  • FEI_front_faces

    Includes 200 different faces shot (Full body shot) from the FEI database.

auxiliary_materials

It includes all the auxiliary materials: reference papers.

Addition

There are many other state-of-art methods to do the same thing. Please refer to the recent best student paper on CVPR about background matting. And I just noticed last week that there are a library called "kornia" that has a good example of doing mapping as well. I am planning to dip into in the future. Well, thanks for visiting my repo. And, happy tiger year! (2022.01.28, xianglun918)