This is the repository for the project - Camera pose tracking using visual odometry
---------- Our Implementation ---------------------------------------- Opencv Implementation -----------
- Python3
- Numpy
- cv2
- math
- glob
- matplotlib
- Clone this repo or extract the "nalindas_proj_5.zip" file.
- Navigate to the folder "Code"
- Inside the python script -
DataPreprocess.py
, you need to specify the appropriate original dataset path in thefile_path
variable. Next, you need to specify the appropriate camera model parameters path in thecamera_path
variable. Finally, you need to specify the path where you want the processed undistorted images to be saved in thewrite_path
variable. - Inside the python script -
main.py
, you need to specify the processed undistorted dataset path which you had specified inDataPreprocess.py
in thepath
variable. Next, you need to specify the appropriate camera model parameters path in thecamera_path
variable. Finally, you need to specify the path where you want the camera pose plot images to be saved in thesave_path
variable. - In the terminal, run the command
python main.py
to run our implementation. Run the commandpython builtin_main.py
to run the opencv implementation. (You need to specify the appropriate original dataset path in theIMAGES_PATH
variable. Finally, you need to specify the path where you want the camera pose plot images to be saved in theMODELS_PATH
variable.)