/GCP_Detection

Automatic Ground Control Point Markers Detection

Primary LanguagePython

#Automatic GCP Marker Detection and locating GCP

#Detailed Approach has been explained in Automatic_GCP_detection_approach.pdf

#Requirement:

  1. python3
  2. pytorch
  3. opencv2
  4. cuda

#to_run_model1 $ python3 main1.py

This performs good with less false predictions. But unable to detect very small GCP markers of size less than 16x16.

#to_run_model2 $ python main2.py

This performs relatively better. But False predictions are more than model 1.

Tests are performed on

  1. CV-Assignment-Dataset
  2. ML - Dataset #2
    (Link to Download https://drive.google.com/drive/folders/1V72hs79YpWrfg3h6s4wYS9Erb0ZdLtEV)

GCP marker are highlighted by Blue rectangle and GCP is located by Red point in the image

Results are located in folder

  1. CV-Assignment-Dataset-Output
  2. CV-Assignment-Dataset-Output-model2
  3. ML - Dataset #2-Ouput