akdhandy/Visual-Odometry-ENPM-673

Visual-Inertial-Odometry

In this project we implement the pipeline for Visual Odometry (VO) from scratch on the Oxford dataset given, and compare it with the implementation using OpenCV built-in functions.

DEPENDENCIES:

• Python Version: 3.x
• Matplotlib.pyplot
• pandas
• xlrd
• OpenCV 3.4 (For the purpose of using SIFT)

Authors

• Arjun Srinivasan
• Arun Kumar
• Praveen

To run the codes:

1.)Run the python files in the current directory which contains all the codes.

2.)Two python scripts namely:

cvvom - which uses inbuilt CV functions to solve the Visual Odometry problem

myvom - which uses our own implementation of functions to solve.

3.)Place the relative path of the images and the model in both the python scripts CV_VOM and myVOM and run: pathimage="/home/arjun/Desktop/VOM/Oxford_dataset/stereo/centre/"

pathmodel="/home/arjun/Desktop/VOM/Oxford_dataset/model/".

In order to get more information visit the below GitHub link: https://github.com/akdhandy/Visual-Odometry-ENPM-673.git