Data types for ORB feature detector

asked 2017-05-16 11:21:02 -0600

UsmanArif gravatar image

Hi, I am receiving images from a ROS Publisher Node. A Subscriber node receives the image, converts it using cvbridge into a Mat object with MONO8 encoding. Then ORB feature detector detects the features in this image. But the problem is the ORB doesn't detect any thing. The same code is working when i read the same image from memory using imread command. I think this is due to MONO8 encoding. So which encoding is compatible with ORB feature detection in Opencv.?

edit retag flag offensive close merge delete

Comments

ORB uses grayscale image for keypoint and descriptors.

What is MONO8 encoding?

LBerger gravatar imageLBerger ( 2017-05-16 11:32:20 -0600 )edit

mono8: CV_8UC1, grayscale image MONO8 is also gray scale encoding, with each pixel of 8bits http://wiki.ros.org/cv_bridge/Tutoria...

UsmanArif gravatar imageUsmanArif ( 2017-05-16 12:20:50 -0600 )edit

save images in both case and compare with a basic software (ms-paint ....)

LBerger gravatar imageLBerger ( 2017-05-16 12:52:11 -0600 )edit
add a comment