AllenCell / AllenCell/allencell-segmenter-ml

Display prediction results correctly for on-screen images

Aperta
#456 0 commenti 0 reazioni 1 assegnatario Rivendicata da @saeliddp Vedi su GitHub
🚨 high priority🚨 bug
Lingua principale
Python
Stelle
5
Fork
3
Metriche di merge delle PR
Nessuna PR unita negli ultimi 30g

Descrizione

Currently when we run prediction on on-screen images we display the result by loading all images from the prediction output path into napari. There is an issue with how we're displaying these results

1. If we predict on a multi-channel image on a channel that is not the first channel , the segmentation shown (which is single channel) is displayed on channel 1, and we cannot overlay this segmentation in napari onto the channel we predicted from. We should extract the channel we predicted on from the raw image, and display the result on top of it.
2. If we predict on multiple images, the raw images are at the bottom of the layer list and all the predictions are at the top of the layer list. It would be better if we displayed the segmentations right on top of the raw image in the layer list, and only display the first two layers (the first raw image, and its segmentation).
3. raw images should be labeled with `[raw]` prepended to the image name in the layer list, and seg should be labeled `[seg]`

Guida per i contributori

Nessuna guida per i contributori indicizzata per questo repository

Valutazione

Questa issue non è ancora stata valutata.

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.