construct a 3D model of an object to detect
hi there, i am a newbie to computer vision application and opencv so kindly be gentle :D my application is to detect a cube in the 3D space and apply visual servoing concepts to change pose of the robot arm to catch the cube from what i read, that i have to supply the model of the cube in the format of xml so the cascade function can detect and track my questions are; 1) how to create xml model for the cube? 2) is what i am thinking is right? 3) is there any possible way to build the model of the cube and how? thanks your time and cooperation is highly appreciated
no, training a cascade on a 3d cube won't work. (it does not have enough features, also it's basically a 2d model, while your cube will show a lot of different poses.)
my cube has actually some hole(circular objects) in it, but till now i can't find a way to be detected and thus followed when i use SURF or SIFT a lot of features are generated with zero control to pick the features attached one me object