Skip to content

Latest commit

 

History

History
38 lines (26 loc) · 1.77 KB

1-get-an-overview.md

File metadata and controls

38 lines (26 loc) · 1.77 KB

Step 1: Get an overview

We will as usual start by presenting an overview of the method and the contents of this project.

Algorithm overview

The main steps in today's lab are:

  • Apply feature detectors and descriptors available in OpenCV to detect and describe keypoints.

    • Detect keypoints in the current frame.
    • Press space to set the current frame as the reference frame.

  • Match keypoints between the reference frame and new current frames using OpenCV, and extract good matches by applying the ratio test.

  • Use the point correspondences to estimate a homography between two images, using RANSAC and the (normalized) DLT:

    • Find a large inlier set by applying a DLT-estimator repeatedly on a minimal set of correspondences using RANSAC.
    • Apply normalized DLT on the inlier set.

  • Use the estimated homography to combine the two images in an image mosaic:

    • Warp a downscaled reference image into the mosaic image.
    • Warp the current frame into the mosaic image based on the estimated homography.

Introduction to the project source files

We have chosen to distribute the code on the following modules:

  • lab_mosaic.py

    Contains the main loop of the program and all exercises. Your task will be to finish the code in this module.

  • common_lab_utils.py

    This module contains utility functions and classes that we will use both in the lab and in the solution. Please take a quick look through the code.

  • solution_mosaic.py

    This is our proposed solution to the lab. Please try to solve the lab with help from others instead of just jumping straight to the solution ;)

    Please continue to the next step to get started with features!