/feature-points

Project detect and matches local features in images. To detect feature, I use harris, log(blob) and dog. To match feature, i use sift, with 3 detector: harris, blob and dog

Primary LanguageC++

|HCMUS|

us579 | Project in Subject Computer Vision

Introduction

This is a C++ implementation of Harris, Blob-LoG, DoG, detection feature point algorithm. And SIFT, matche feature point.

Libraries used

opencv I bulid algorithms with C++ and opencv You can setup opencv in Visual Studio here: https://www.youtube.com/watch?v=unSce_GPwto&pp=ygUcdXNlIG9wZW5jdiBpbiB2aXN1YWwgc3R1ZGlvIA%3D%3D

Usage example

Find feature point and show result:

Use harris

feature_points.exe <path_image> harris

Result: Matching result Use blob

feature_points.exe <path_image> blob

Result: Matching result Use dog

feature_points.exe <path_image> dob

Result: Matching result Match features in two images and show the result:: Use SIFT

feature_points.exe <path_image_1> <path_image_2> sif <detector>

Parameter :

  • 1: use harris
  • 2: use blob
  • 3: use dog Result: Matching result

You can see another example image here

Useful links

  • You can find more information about them in report.pdf