Re^3: Abstract image registration or feature detection

Weird, perl 5.30.0 with PDL 2.019 (which is the combo that shipped with Strawberry Perl's PDL edition) doesn't give that error.

These are the outputs for my run (adding a print "perl version $]";print "PDL version $PDL::VERSION"; at the beginning to print the versions):

perl version 5.030000
PDL version 2.019

[
 [ 58 108 186 255 331]
 [ 48 155  80 191  48]
 [  1   1   1   1   1]
]


[
 [471 531 603 682 747]
 [ 15 141  90 227 107]
 [  1   1   1   1   1]
]


[
 [125 197 282]
 [ 73 158  94]
 [  1   1   1]
]


[
 [542 622 701]
 [ 64 175 138]
 [  1   1   1]
]

938
522
5
224210
71514
98997
3034
580
617991
126140
319271
78268

[
 [224210  98997    938]
 [ 98997  71514    522]
 [   938    522      5]
]


[
 [617991 126140]
 [319271  78268]
 [  3034    580]
]


[
 [  1.0099669  0.33666059]
 [0.084675354   1.0198999]
 [  408.49011  -53.635077]
]


[
 [  1.0099669 0.084675354   408.49011]
 [ 0.33666059   1.0198999  -53.635077]
 [          0           0           1]
]


[
 [ 540.91727  620.83229  701.26025]
 [  62.90019  173.83124   137.1738]
 [         1          1          1]
]


[
 [ -1.0827329  -1.1677124   0.2602499]
 [   -1.09981   -1.168756 -0.82619875]
 [          0           0           0]
]
[download]

Mathematically speaking, a [3x3] x [3x2] is the right balance of dimensions for a successful matrix multiplication (resulting in a [3x2] matrix). The use of PDL::Matrix used to be the way to make PDL match mathematical order for the dimensions, but maybe they've changed that between 2.019 and 2.057. (I don't think I can test, because the Strawberry Perl PDL setup is rather dependent on the libraries that Strawberry bundles, and I don't know if upgrading PDL will work... maybe I'll try in a sandbox to avoid ruining my "main" installation).

I tried getting rid of the use PDL::Matrix, but then that gets rid of the mpdl command, causing errors during perl's compile stage, and I'm not sure what hoops I would need to jump through to make my script compatible with both 2.019 and 2.057.

Comment on Re^3: Abstract image registration or feature detection Select or Download Code

Replies are listed 'Best First'.
Re^4: Abstract image registration or feature detection by pryrt (Abbot) on Jul 04, 2022 at 20:54 UTC
Okay, 5.30.2 & PDL 2.021 and 5.32.1 & PDL 2.025 (the last pre-bundled Perl+PDL from Strawberry) both worked like my 5.30.0 & PDL 2.019. But I was able to install PDL 5.080 with Perl 5.32.1, to get slightly newer PDL but slightly older Perl than you. But it showed the same dimensional difference. So I did some debug to make it work without PDL::Matrix #!perl -l use 5.012; # //, strict, say use warnings; use PDL 2.057; #use PDL::Matrix; print "perl version $]"; print "PDL version $PDL::VERSION"; print my $red_in = (pdl [ # reference [419 , 17 ], # p01 [772 , 13 ], # p02 [348 , 102], # p03 [720 , 95 ], # p04 [58 , 238], # p05 [218 , 236], # p06 [55 , 333], # p07 [58 , 391], # p08 [60 , 498], # p09 [58 , 554], # p10 [1061, 553], # p11 [60 , 892], # p12 ])->append(ones(1))->transpose; print my $red_out = (pdl [ # target_41 [487, 466], # p01 [531, 630], # p02 [439, 447], # p03 [486, 618], # p04 [340, 328], # p05 [358, 401], # p06 [297, 343], # p07 [274, 349], # p08 [227, 366], # p09 [205, 374], # p10 [313, 809], # p11 [77 , 420], # p12 ])->append(ones(1))->transpose; print my $blue_in = (pdl [ # reference [438,331], # q01 [432,392], # q02 [436,493], # q03 [436,552], # q04 [439,602], # q05 ])->append(ones(1))->transpose; print my $blue_out_for_checking = (pdl [ # target_41 [339,508], # q01 [314,518], # q02 [265,533], # q03 [245,537], # q04 [228,544], # q05 ])->append(ones(1))->transpose; $\ = ''; print for "sx = ", my $sx = sum( $red_in->transpose->slice(0) ), + "\n", "sy = ", my $sy = sum( $red_in->transpose->slice(1) ), + "\n", "s1 = ", my $s1 = sum( $red_in->transpose->slice(2) ), + "\n", "sxx = ", my $sxx = sum( $red_in->transpose->slice(0)2 ), + "\n", "syy = ", my $syy = sum( $red_in->transpose->slice(1)2 ), + "\n", "sxy = ", my $sxy = sum( $red_in->transpose->slice(0)$red_in->tra +nspose->slice(1) ), "\n", "sv = ", my $sv = sum( $red_out->transpose->slice(0) ), + "\n", "sw = ", my $sw = sum( $red_out->transpose->slice(1) ), + "\n", "sxv = ", my $sxv = sum( $red_in->transpose->slice(0)$red_out->tr +anspose->slice(0) ), "\n", "sxw = ", my $sxw = sum( $red_in->transpose->slice(0)$red_out->tr +anspose->slice(1) ), "\n", "syv = ", my $syv = sum( $red_in->transpose->slice(1)$red_out->tr +anspose->slice(0) ), "\n", "syw = ", my $syw = sum( $red_in->transpose->slice(1)*$red_out->tr +anspose->slice(1) ), "\n", ; $\ = "\n"; print "G => ", my $G = pdl [ [ $sxx, $sxy, $sx ], [ $sxy, $syy, $sy ], [ $sx, $sy, $s1 ], ]; print "Q => ", my $Q = pdl [ [ $sxv, $sxw ], [ $syv, $syw ], [ $sv, $sw ], ]; print "P => ", my $P = inv($G) x $Q; # print "P->transpose", $P->transpose; # print "P->append->transpose", $P->append(zeros(1))->transpose; print "abcdef001 => ", my $abcdef001 = $P->append(zeros(1))->transpose +->set(2,2,1); print "blue out calc => ", my $blue_out_calc = $abcdef001 x $blue_in; print "error => ", $blue_out_calc - $blue_out_for_checking; # no more +than one pixel off __END__ perl version 5.032001 PDL version 2.080 [ [ 419 772 348 720 58 218 55 58 60 58 1061 60] [ 17 13 102 95 238 236 333 391 498 554 553 892] [ 1 1 1 1 1 1 1 1 1 1 1 1] ] [ [487 531 439 486 340 358 297 274 227 205 313 77] [466 630 447 618 328 401 343 349 366 374 809 420] [ 1 1 1 1 1 1 1 1 1 1 1 1] ] [ [438 432 436 436 439] [331 392 493 552 602] [ 1 1 1 1 1] ] [ [339 314 265 245 228] [508 518 533 537 544] [ 1 1 1 1 1] ] sx = 3887 sy = 3922 s1 = 12 sxx = 2604611 syy = 2052390 sxy = 929565 sv = 4034 sw = 5551 sxv = 1608891 sxw = 2354880 syv = 945962 syw = 1755275 G => [ [2604611 929565 3887] [ 929565 2052390 3922] [ 3887 3922 12] ] Q => [ [1608891 2354880] [ 945962 1755275] [ 4034 5551] ] P => [ [0.11504241 0.44420653] [-0.4325134 0.11994661] [ 440.26231 279.49489] ] abcdef001 => [ [0.11504241 -0.4325134 440.26231] [0.44420653 0.11994661 279.49489] [ 0 0 1] ] blue out calc => [ [ 347.48895 320.41538 277.19169 251.6734 230.39286] [ 513.75967 518.41118 532.30261 539.37946 546.70941] [ 1 1 1 1 1] ] error => [ [ 8.488949 6.4153774 12.191694 6.6734037 2.3928612] [ 5.7596723 0.41117608 -0.69739064 2.3794591 2.709409] [ 0 0 0 0 0] ] [download] This shows the 12-ish pixel error, but will hopefully work with your 2.057 (and definitely worked with my 2.080).	[reply] [d/l]
Re^4: Abstract image registration or feature detection by kikuchiyo (Hermit) on Jul 04, 2022 at 21:01 UTC
Minimal pair: `perl -MPDL -MPDL::Matrix -e '$m = mpdl [[2,0],[0,2]]; $v = vpdl [1,2]; + print $m x $v' [ [2] [4] ]` [download] vs. `perl -MPDL -MPDL::Matrix -e '$m = mpdl [[2,0],[0,2]]; $v = vpdl [1,2]; + print inv($m) x $v' Dim mismatch in matmult of [2x2] x [2x1]: 2 != 1 at /usr/lib64/perl5/v +endor_perl/PDL/Primitive.pm line 274. PDL::matmult(PDL=SCALAR(0x5600a82130e0), PDL::Matrix=SCALAR(0x5600 +a8202ba0), PDL=SCALAR(0x5600a8d718c8)) called at /usr/lib64/perl5/ven +dor_perl/PDL/Primitive.pm line 31 PDL::__ANON__(PDL=SCALAR(0x5600a82130e0), PDL::Matrix=SCALAR(0x560 +0a8202ba0), "") called at -e line 1` [download] This looks like a bug in PDL::Matrix, specifically with inv.	[reply] [d/l] [select]
Re^5: Abstract image registration or feature detection by etj (Priest) on Jul 05, 2022 at 20:51 UTC
As now noted on pryrt's excellent bug report (thank you), this will still work as expected if you switch your `inv($m)` to `$m->inv`. This will then engage PDL::Matrix's overload of `x`, which is what you need here. I would strongly encourage you not to use PDL::Matrix, because somewhat similar to PDL::Complex, it is full of gotchas. Even though after some head-scratching it became clear what exactly was happening here, it's not particularly clear how to fix it - a keyhole change to the Perl function PDL::MatrixOps::inv to make it ensure the output was the right class would resolve this exact issue. But since that uses lu_backsub for its work, and that is an eye-wateringly complicated Perl function as well, it will not be easy. That function's output variable comes from some PDL::Slices operations. If someone were minded to at least use `git bisect` to identify the commit that broke it, that would be very helpful. Even better would be a pull-request that fixes this (obviously with a test, in `t/matrix.t`), that will use our amazing CI to check it doesn't break anything else in main PDL or known downstreams. A top tip would be that `make coretest` runs `t/matrix.t`, with a very minimal build otherwise that allows a very quick dev cycle. As a hopefully practical (as well as better-performance, because it uses LAPACK) alternative, have you taken a look at PDL::LinearAlgebra?	[reply] [d/l] [select]
Re^5: Abstract image registration or feature detection by pryrt (Abbot) on Jul 05, 2022 at 19:28 UTC
kikichiyo, Thanks for that SSCCE. I included it in a bug report	[reply]
Re^4: Abstract image registration or feature detection by etj (Priest) on Jul 05, 2022 at 21:26 UTC
The way to avoid needing `mpdl` is to transpose your data, and use `pdl`, and I would recommend doing so. But as observed elsewhere, there are almost certainly out-of-the-box solutions to this particular problem.	[reply] [d/l] [select]