fix: image creator function

[jkbmrz]

Purpose

Outputs of the ImageOutputParser are not visualized correctly in DAI Visualizer due to a bug in the create_image_message() creator.

Specification

Changes in create_image_message():

• Use .setCvFrame() instead of .setData() when creating the dai.ImgFrame,
• add an argument (img_frame_type) to adjust the output ImgFrame.Type to the running platform.

Changes in ImageOutputParser:

• extract running platform and pass the correct ImgFrame.Type when calling the create_image_message().

Dependencies & Potential Impact

None

Deployment Plan

None

Testing & Validation

The changes were tested by running the generic-example both on RVC2 and RVC4 for the following models:

• rt-super-resolution
• esrgan (RVC4 only)
• zero-dce

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 75.00000% with 1 line in your changes missing coverage. Please review.

Project coverage is 47.50%. Comparing base (69ffd94) to head (7f8a292).
Report is 1 commits behind head on main.

✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
depthai_nodes/node/parsers/image_output.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #188      +/-   ##
==========================================
+ Coverage   41.63%   47.50%   +5.86%     
==========================================
  Files          75       75              
  Lines        4285     4286       +1     
==========================================
+ Hits         1784     2036     +252     
+ Misses       2501     2250     -251     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[jkbmrz]

To discuss: ImageOutputParser currently always ouputs dai.ImgFrame of type BGR888i. This works for the simple visualizations (e.g. generic-example) but might we want to adjust it to the device the pipeline is running on (e.g. if input to the model is BGR888p, that the parser also outputs BGR888p instead of BGR888i as is the case now)?

[kkeroo]

but might we want to adjust it to the device the pipeline is running on

We can since it is probably easy addition (self.getParentPipeline()...). But the change will probably be useful only if we send the message to the new NN node because in the visualization you can get the frame out with getCvFrame()