[INSTRM-2157] Same spot is reported twice in mcs_data Created: 25/Jan/24  Updated: 07/Feb/24

Status: Open
Project: Instrument control development
Component/s: ics_mcsActor
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Normal
Reporter: arnaud.lefur Assignee: karr
Resolution: Unresolved Votes: 0
Labels: EngRun
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified


 Description   

I have many many example of this bug where the same mcs spot end-up being reported twice in the database with slight shift in the centroid, usually 0.1 pixel.

I think that's up to the centroid algorithm to handle this properly.

Here are few examples but there are many more.

10345300 spotId(379) mcsCenter(x,y) 5771.38965,1401.47644
10345300 spotId(377) mcsCenter(x,y) 5771.39355,1401.47571
10345900 spotId(2157) mcsCenter(x,y) 2398.24268,3821.18799
10345900 spotId(2156) mcsCenter(x,y) 2398.31738,3821.1936
10345900 spotId(1071) mcsCenter(x,y) 2816.32617,3004.79907
10345900 spotId(1067) mcsCenter(x,y) 2816.36719,3004.76831
10345900 spotId(304) mcsCenter(x,y) 2514.52271,1873.76746
10345900 spotId(303) mcsCenter(x,y) 2514.57446,1873.65552
10345900 spotId(41) mcsCenter(x,y) 4217.8291,1295.82141
10345900 spotId(37) mcsCenter(x,y) 4218.09277,1295.67822
10346300 spotId(1974) mcsCenter(x,y) 3685.72192,4180.08057
10346300 spotId(1973) mcsCenter(x,y) 3685.729,4180.07959
10346500 spotId(1974) mcsCenter(x,y) 3685.93457,4180.0498
10346500 spotId(1973) mcsCenter(x,y) 3685.96826,4180.04932
10346500 spotId(308) mcsCenter(x,y) 6394.00146,1590.0553
10346500 spotId(307) mcsCenter(x,y) 6394.00488,1590.05554
10346700 spotId(308) mcsCenter(x,y) 6394.14697,1589.82617
10346700 spotId(307) mcsCenter(x,y) 6394.16113,1589.82568
10350400 spotId(1278) mcsCenter(x,y) 3687.84473,4179.03662
10350400 spotId(1277) mcsCenter(x,y) 3687.8606,4179.03613
10350400 spotId(1068) mcsCenter(x,y) 5039.84521,3435.86133
10350400 spotId(1067) mcsCenter(x,y) 5039.8501,3435.86182
10350400 spotId(193) mcsCenter(x,y) 6395.68359,1588.88428
10350400 spotId(192) mcsCenter(x,y) 6395.74023,1588.87952
10350700 spotId(307) mcsCenter(x,y) 6395.69971,1588.92615
10350700 spotId(306) mcsCenter(x,y) 6395.6958,1588.89221

10439000 spotId(157) mcsCenter(x,y) 3690.28174,4177.67578
10439000 spotId(27) mcsCenter(x,y) 6398.04053,1587.69055
10439000 spotId(26) mcsCenter(x,y) 6398.08252,1587.66919
10443800 spotId(153) mcsCenter(x,y) 3690.49878,4177.68701
10443800 spotId(152) mcsCenter(x,y) 3690.51123,4177.68652
10449200 spotId(448) mcsCenter(x,y) 5196.2041,317.89389
10449200 spotId(444) mcsCenter(x,y) 5196.19971,317.893677
10449200 spotId(674) mcsCenter(x,y) 4223.40527,3749.88477
10449200 spotId(671) mcsCenter(x,y) 4223.45508,3749.86841
10451800 spotId(153) mcsCenter(x,y) 3687.00342,4179.71729
10451800 spotId(152) mcsCenter(x,y) 3687.01367,4179.71191
10451800 spotId(26) mcsCenter(x,y) 6394.854,1589.64294
10451800 spotId(25) mcsCenter(x,y) 6394.89062,1589.65698
10453600 spotId(1223) mcsCenter(x,y) 6626.8291,2469.39185
10453600 spotId(1214) mcsCenter(x,y) 6626.82861,2469.39185
10456900 spotId(154) mcsCenter(x,y) 3687.49023,4177.75195
10456900 spotId(153) mcsCenter(x,y) 3687.50024,4177.70703
10458700 spotId(1224) mcsCenter(x,y) 6627.40723,2467.34448
10458700 spotId(1215) mcsCenter(x,y) 6627.42432,2467.34253
10458700 spotId(207) mcsCenter(x,y) 3233.06616,1407.81201
10458700 spotId(202) mcsCenter(x,y) 3233.08496,1407.80298
10461700 spotId(156) mcsCenter(x,y) 3688.19727,4177.62061
10461700 spotId(155) mcsCenter(x,y) 3688.25439,4177.61182
10461700 spotId(28) mcsCenter(x,y) 6395.94971,1587.58887
10461700 spotId(27) mcsCenter(x,y) 6396.09473,1587.5824
10462500 spotId(748) mcsCenter(x,y) 4534.49512,4961.11182
10462500 spotId(746) mcsCenter(x,y) 4534.49658,4961.11182
10462500 spotId(64) mcsCenter(x,y) 3233.58789,1407.9762
10462500 spotId(61) mcsCenter(x,y) 3233.60352,1407.97412
10467800 spotId(155) mcsCenter(x,y) 3688.92676,4177.67139
10467800 spotId(154) mcsCenter(x,y) 3688.94287,4177.71826
10472600 spotId(159) mcsCenter(x,y) 3689.5769,4177.5166
10472600 spotId(158) mcsCenter(x,y) 3689.6062,4177.51611
10472700 spotId(1816) mcsCenter(x,y) 2403.06519,3824.48047
10472700 spotId(1815) mcsCenter(x,y) 2403.05444,3824.48096
10475300 spotId(154) mcsCenter(x,y) 3689.65771,4177.60986
10475300 spotId(153) mcsCenter(x,y) 3689.69312,4177.6084



 Comments   
Comment by arnaud.lefur [ 25/Jan/24 ]

is it a consequence of MCS being badly focused ? I know some refocusing was done but I don't know what was the visit range.

Comment by yuki.moritani [ 25/Jan/24 ]

MCS was defocused during the run 14 entirely.. (say, visit =103327 – 105058). Is this from mcs_data table?

Comment by karr [ 25/Jan/24 ]

Thanks for the report; I'll take a look at the data and see what's happening.

Comment by karr [ 06/Feb/24 ]

I took a look at the first image, 10345300, and ran it through my local copy of the code, and was not able to reproduce the error. I detect all the same spots, except for the two near duplicates. I'm working from a copy of the master branch, and recompiled the centroid code. 

Is there any way to check what branch/version of the actor was being used when the data were taken? 

Comment by yuki.moritani [ 06/Feb/24 ]

It is strange... According to the actor log, it says

2023-12-22 14:54:22.994Z cmds 20 CommandLink.py:122 > 2 2 i version="1.8.7-25-gc6c0064-dirty"

(I grepped the log for the frameid) But it should be the same as "tag 1.8.8". chyan Could you confirm it?

And just I also record here that the duplication in centroid happened in July too. 

Comment by chyan [ 07/Feb/24 ]

The difference between 1.8.8 and the running developing branch is the image saving method. There is no difference in centroiding method.

Generated at Sat Feb 10 16:43:34 JST 2024 using Jira 8.3.4#803005-sha1:1f96e09b3c60279a408a2ae47be3c745f571388b.