feat: run detection

src/aind_exaspim_soma_detection/run_detection.py

@@ -0,0 +1,61 @@
+"""
+Created on Wed Nov 27 12:00:00 2024
+
+@author: Anna Grim
+@email: [email protected]
+
+Code that generates soma proposals.
+
+"""
+
+from random import sample
+from time import time
+
+import warnings
+
+from aind_exaspim_soma_detection import soma_proposal_generation as spg
+from aind_exaspim_soma_detection.utils import img_util, util
+
+warnings.filterwarnings("ignore", category=RuntimeWarning)
+
+
+def run():
+    # Proposal Generation
+    t0 = time()
+    print("\nBrain_ID:", brain_id)
+    print("\nStep 1: Soma Proposal Generation")
+    proposals = spg.run_on_whole_brain(
+        img_prefix,
+        overlap,
+        patch_shape,
+        multiscale,
+        bright_threshold=bright_threshold,
+    )
+    print("\n# Proposals Generated:", len(proposals))
+    print("Runtime:", time() - t0)
+
+    # Save Results
+    output_dir = "/root/capsule/results/soma_proposals"
+    util.write_points(output_dir, proposals, color="1.0 0.0 0.0", prefix="proposal_")
+
+    proposals = sample(proposals, 200)
+    output_dir = "/root/capsule/results/soma_proposals_training"
+    util.write_points(output_dir, proposals, color="1.0 0.0 0.0", prefix="proposal_")
+
+
+if __name__ == "__main__":
+    # Parameters
+    bucket_name = "aind-open-data"
+    brain_id = "704522"
+    multiscale = 4
+
+    bright_threshold = 10
+    overlap = (28, 28, 28)
+    patch_shape = (64, 64, 64)
+
+    # Initializations
+    img_prefixes = util.read_json("/root/capsule/data/exaspim_image_prefixes.json")
+    img_prefix = img_prefixes[brain_id] + str(multiscale)
+
+    # Main
+    run()