Can Orbit save the image on GUI after object detection?

Hello,

After object detection the GUI gives a stats about how many classified object.

However, I need to do some non-ridge image registration and down sampling of the image after the object detection. Thus, it would be convenient if I can save the resulting image in GUI as a binary mask (or three/four binary color map), such as the last picture in the tutorial (https://www.orbit.bio/help/object-classification/) .

The current Tool-save classified image or the script only saves the image after color-based classification (no object detection overlay or color-coded image)

(https://github.com/mstritt/orbit-image-analysis/blob/22462cc896f8089277cb7932cdf9e2469168901d/src/main/groovy/com/actelion/research/orbit/imageAnalysis/scripts/SaveClassificationImage.groovy)

Thanks!

Hi @jena,

well we can do this, however, I guess what you really want is to get the outlines and labels of the segmented objects.
From the SegmentationResult segmentationResult (which you get from OrbitHelper.Segmentation(…)) you can retrieve all outlines from segmentationResult.getShapeList().
Then you can use these Shapes for applying your non-rigid registration…

Regards,
Manuel

1 Like

Thank you very much.

Could I do the save GUI image method first? It might be easier for me to get a first glance.
I will try the SegmentationResult then.

Thanks!

you can only save the classification map easily. To save the rendered “color labeled” objects you to have to look into the paintComponent() method of RecognitionFrame where the rendering happens.
I recommend to use the shapes and either render them on your own to the image or extract the shape coordinates and do something with it.
Regards,
Manuel