OpenCV SURF Image Matching on Android Phone

OpenCV SURF Image Matching: Find same points between two images.

An image matching demo that uses OpenCV SURF (Speeded Up Robust Features) descriptors. Images can be selected from phone's gallery.

The find_obj.cpp demo in OpenCV samples is slightly modified and used. Image matching of default images took 5.162 sec on my Samsung Galaxy S phone. Size of images is 320*240.

Download here
https://market.android.com/details?id=test.photo3dapps.surf&feature=searc...

For more info
http://photo3d-apps.blogspot.com/

Category:

Science & Technology

Tags:

• SURF
• OpenCV
• Android
• Galaxy S
• image matching
• image processing

[osamabinladin4u]

when i try to use a picture from my album it the app gives an error saying that i must only use images less 3.4 Mpixels . Is there a way to fix this? thanks

[photo3dapps]

@osamabinladin4u Hi, since limited memory and CPU processing power, I restricted picture size. I can reduce the size of picture larger than 3.4MP before processing; however, at this time, I don't have plan to do it. Thanks :)

[osamabinladin4u]

@photo3dapps Is it possible if yo can provide the code? Thanks in advance.

[photo3dapps]

@osamabinladin4u No, sorry.

[luiskov]

thumbs up for that varitech fighter pic you used.

[photo3dapps]

@luiskov Thanks, it's my one of favorites

[TheMegusAlexandrus]

@fragfoo I think its fine, as the computation is being done on the Android platform itself. So 5 seconds is not very bad. SURF matching is also done in many augmented reality apps, and they too have a delay of about 3 secs.

[fragfoo]

SURF is supposed to match features near real-time so your 5sec is uber slow.