nmellado / Super4PCS

Efficient Global Point-cloud registration
http://nmellado.github.io/Super4PCS/
Other
455 stars 215 forks source link

Different result of your video demonstration #70

Closed sebastien247 closed 3 years ago

sebastien247 commented 4 years ago

Hi, First, thanks for your work. I implemented OpenGR with the VTK render to see the result. Unfortunately, I didn't obtain the same result as this video: https://www.youtube.com/watch?v=TgHrotscPCA whether on Super4PCS or 4PCS.

this is options applied for Super4PCS and 4PCS (same options in run-example.bat):

    OptionType options;
    options.configureOverlap( 0.7 );
    options.delta = 0.01;
    options.sample_size = 200;
    options.max_time_seconds = 1000;

It's the result matrix:

  | [0] | 0.78860002756118774 | double
  | [1] | 0.23901370167732239 | double
  | [2] | 0.56655317544937134 | double
  | [3] | 0.00000000000000000 | double
  | [4] | -0.10672703385353088 | double
  | [5] | 0.96058332920074463 | double
  | [6] | -0.25668844580650330 | double
  | [7] | 0.00000000000000000 | double
  | [8] | -0.60557353496551514 | double
  | [9] | 0.14195793867111206 | double
  | [10] | 0.78302532434463501 | double
  | [11] | 0.00000000000000000 | double
  | [12] | -0.13029700517654419 | double
  | [13] | 0.024866815656423569 | double
  | [14] | -0.025774050503969193 | double
  | [15] | 1.0000000000000000 | double

image

Thanks for your help

Sébastien

nmellado commented 3 years ago

Hi, Super4PCS library is now deprecated and superseded by https://github.com/STORM-IRIT/OpenGR Could you please retry your experiment on OpenGR master branch, open a dedicated issue there, share the dataset used, and detail your compiler and operating system. Cheers Nicolas