Execution in non-interactive environment about segment-geospatial HOT 3 CLOSED

Hi @lymperis-e , I think your problem is relative to QGIS and how you call pytorch in the background, not segment-geospatial.

To avoid freeze and errors in one of my script, I had to add this code to the class definition :

class Process(QgsProcessingAlgorithm):
    ...
    def flags(self):
        return super().flags() | QgsProcessingAlgorithm.FlagNoThreading

Not sure if it's the same logic in plugins...
But I'm pretty sure this is a threading problem, which is common when coding PyQt (except if it is indeed just a slow CPU).

Hope this will help.

(Edit : it does not look like there are similar flags for plugins, but may be wrapping your process in a QgsProcessingAlgorithm will allow you to better control its execution).

from segment-geospatial.

It works fine on my end. Do you have a GPU? Running on CPU can be pretty slow and might look like frozen.

import os
import torch
from samgeo import SamGeo, tms_to_geotiff

bbox = [-95.3704, 29.6762, -95.368, 29.6775]
image = 'satellite.tif'
tms_to_geotiff(output=image, bbox=bbox, zoom=20, source='Satellite')

out_dir = os.path.join(os.path.expanduser('~'), 'Downloads')
checkpoint = os.path.join(out_dir, 'sam_vit_h_4b8939.pth')

device = 'cuda' if torch.cuda.is_available() else 'cpu'
sam = SamGeo(
    checkpoint=checkpoint,
    model_type='vit_h',
    device=device,
    erosion_kernel=(3, 3),
    mask_multiplier=255,
    sam_kwargs=None,
)

mask = 'segment.tiff'
sam.generate(image, mask)

vector = 'segment.gpkg'
sam.tiff_to_gpkg(mask, vector, simplify_tolerance=None)

from segment-geospatial.

@vidlb Thank you for helping out. Closing this issue for now. Feel free to reopen it if you are still experiencing issues @lymperis-e

from segment-geospatial.