search

tum-vision / lsd_slam

LSD-SLAM
GNU General Public License v3.0
2.59k stars 1.23k forks source link

how use variable baseline to calculate inverseDepth #203

Closed tsingjinyun closed 8 years ago

tsingjinyun commented 8 years ago

hi,everyone: in the .../src/DepthEstimation/Depth.cpp, in doLineStereo() function, https://github.com/tum-vision/lsd_slam/blob/master/lsd_slam_core/src/DepthEstimation/DepthMap.cpp#L1874 the author use idnew_best_match = (dot0 - oldX*dot2) / nominator; and i think this used 1/z=d/fB (where f is the focal length,B is the baseline,z is the pixel depth),then the question come here: i find it hard to match this formula. anybody come across this problem, thanks for your explanation

tsingjinyun commented 8 years ago

@JakobEngel，the alpha is the length of the searched inverse depth interval and the length of the searched epipolar line segment ratio， https://github.com/tum-vision/lsd_slam/blob/master/lsd_slam_core/src/DepthEstimation/DepthMap.cpp#L1896 and those lines codes is very important to calculate depth and variance, any mathematics inference behind it?

HeYijia commented 8 years ago

Hi @tsingjinyun , I think I find it out. It can be derived from the formula of the projection of a 3D world point into a camera image. If have any question, let me know, GOOD LUCK. qq 20160818094752

tsingjinyun commented 8 years ago

Thanks, I have figure it out. Thanks,anyway

发自 网易邮箱大师 On 08/18/2016 09:52, HeYijia wrote:

Hi @tsingjinyun , I think I find it out. It can be derived from the formula of the projection of a 3D world point into a camera image. If have any question, let me know, GOOD LUCK.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.