joycenerd / Computer_Vision_2021

Codes for Computer Vision course in the spring semester at NCTU

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

NCTU-CV_2021

Codes for Computer Vision course in the spring semester at NCTU

Camera Calibration (link)

Camera calibration without using cv2.calibrateCamera()

Image Resolution (link)

Hybrid image, image pyramid, colorizing the Russian empire

Panoramic Stitching (link)

Combining multiple photographic images with overlapping fields of view to produce a segmented panorama

Structure From Motion (link)

Estimating three-dimensional structures from two-dimensional image sequences coupled with local motion signals.

Scene Classification (link)

  • EfficientNet accuracy: 93.78%
  • ResNest accuracy: 93.11%
  • Simple CNN accuracy: 74%
  • BOS SVM accuracy: 68%
  • BOS KNN accuracy: 57.3%
  • Tiny KNN accuracy: 22.67%

About

Codes for Computer Vision course in the spring semester at NCTU

License:MIT License


Languages

Language:Jupyter Notebook 98.6%Language:Python 1.4%Language:MATLAB 0.0%