Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
A Digital Scene Matcher and Area Co-relator using SIFT
C++
tree: 1f4de12e33

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.
results
test-images
CMakeLists.txt
README
SIFT_Descriptor.cpp

README

********************
* SIFT BASED DSMAC *
********************

README:


This code is a SIFT based trackers for UAV applications. The tracker is meant for DMSAC purposes

DSMAC: Digital Scene Matcher and Area Co-relator

What it tries to do is to match a given input from a UAV and match it to a large map.

Since the input object is highly scaled out in the map scene. It is necessary to use PryDown functions to make it easier for them to match.

Certain other boosts in the SIFT codes makes the performance in this DMSAC to around 30ms. The upgrades are already pulled into the new OpenCV 2.4.2
 (For reference, this pulls involved changes in the Lehviestien distance measurement and the feature calculations)

REQUIREMENTS:

OpenCV 2.4.2
CMake 2.8

BUILDIING FROM SOURCE:

1. USE CMAKE

mkdir build
cd build
cmake ..
make

This should make the required files and the executable as well


RUNNING THE CODE:

./SIFT_descriptor <Object image> <Scene Image> <resize parameter in power of two> <number of matches to be considered>

img1 : Object Image (This will be the zoomed in version)
img2 : Scene Image (This will be the Map version, scaled out as compared to the object image)
resize parameter : This parameter will scale the image. and float value k will change the image size k times. In most cases k < 0 (In an UAV, this parameter will be a function of height of the UAV)
number of matches to be considered : This parameter defines the number of matches to be used for the purpose of homography calculations.


AUTHORS:

Abhinav Gupta

CONSTRBUTORS:

Ritesh Ranjan
Something went wrong with that request. Please try again.