Describe the bug
EN
When using a 4-step pipeline, the labels do not receive an ID, which means it cannot be used for alignment or scale.
There is an assumption that the same will happen with any tags, regardless of what kind of tag it is.
RU
При использовании 4 шагового пайплайна, метки не получают ID в результате, чего его не выйдет использовать для выравнивания или приминения масштаба.
Есть предположение, что аналогичное будет с любыми тегами, независимо от того, что это за тег.
EN
In this case, by using the SFMColorizing node, the labels receive a color that replaces their ID, and their ID as a result becomes almost random.
RU
При этом, применяя нод SFMColorizing метки получают цвет, что заменяет их ID и их ID в результате становится около случайным.
To Reproduce
Steps to reproduce the behavior:
- Enable "Photogrammetry Experimental"
- Instead of the label ID, we will get that all labels are equal to 255, since they will receive white color.
Expected behavior
Obtain a working tag ID for further use.
Log
If applicable, copy paste the relevant log output (please embed the text in a markdown code tag "```" )
Without SFMColorizing
* landmarksDescriberTypes = "cctag3"
* maxCoresAvailable = Unknown Type "unsigned int" (default)
* maxMemoryAvailable = 18446744073709551615 (default)
* objectType = Unknown Type "enum EObject"
* verboseLevel = "info"
[18:26:09.058123][info] Hardware:
[18:26:09.058123][info] Detected core count: 16
[18:26:09.058123][info] OpenMP will use 16 cores
[18:26:09.058123][info] Detected available memory: 4012 Mo
[18:26:09.058123][info]
== Landmarks ==
With SFMColorizing
* landmarksDescriberTypes = "cctag3"
* maxCoresAvailable = Unknown Type "unsigned int" (default)
* maxMemoryAvailable = 18446744073709551615 (default)
* objectType = Unknown Type "enum EObject"
* verboseLevel = "info"
[18:36:38.279999][info] Hardware:
[18:36:38.279999][info] Detected core count: 16
[18:36:38.279999][info] OpenMP will use 16 cores
[18:36:38.281060][info] Detected available memory: 7518 Mo
[18:36:38.281060][info]
== Landmarks ==
140 <=> 141: 0.00988576
140 <=> 151: 0.0114316
140 <=> 155: 0.00584645
141 <=> 140: 0.00988576
141 <=> 151: 0.0056986
141 <=> 155: 0.0113591
151 <=> 140: 0.0114316
151 <=> 141: 0.0056986
151 <=> 155: 0.00976553
155 <=> 140: 0.00584645
155 <=> 141: 0.0113591
155 <=> 151: 0.00976553
Desktop (please complete the following and other pertinent information):
- OS: [win 11]
- Python version [*3.14.4, 3.13.13, 3.12.10, 3.9.13]
- Qt/PySide version [e.g. 6.8.2]
- Meshroom version: please specify if you are using a release version or your own build
- Binary version (if applicable) [2025.1.0]
Describe the bug
EN
When using a 4-step pipeline, the labels do not receive an ID, which means it cannot be used for alignment or scale.
There is an assumption that the same will happen with any tags, regardless of what kind of tag it is.
RU
При использовании 4 шагового пайплайна, метки не получают ID в результате, чего его не выйдет использовать для выравнивания или приминения масштаба.
Есть предположение, что аналогичное будет с любыми тегами, независимо от того, что это за тег.
EN
In this case, by using the SFMColorizing node, the labels receive a color that replaces their ID, and their ID as a result becomes almost random.
RU
При этом, применяя нод SFMColorizing метки получают цвет, что заменяет их ID и их ID в результате становится около случайным.
To Reproduce
Steps to reproduce the behavior:
Expected behavior
Obtain a working tag ID for further use.
Log
If applicable, copy paste the relevant log output (please embed the text in a markdown code tag "```" )
Without SFMColorizing
With SFMColorizing
Desktop (please complete the following and other pertinent information):