Got it to successfully create individual bounding boxes for a whole mask

app.py

@@ -6,7 +6,7 @@ import torch
 import pathlib
 from PIL import Image
 
-from transformers import DetrFeatureExtractor, DetrForSegmentation, MaskFormerFeatureExtractor, MaskFormerForInstanceSegmentation
+from transformers import DetrFeatureExtractor, DetrForSegmentation, MaskFormerImageProcessor, MaskFormerForInstanceSegmentation
 from transformers.models.detr.feature_extraction_detr import rgb_to_id
 
 
@@ -73,15 +73,15 @@ def segment_images(model_name,url_input,image_input,threshold):
         pass
     elif "maskformer" in model_name.lower():
         # Load the processor and model
-        processor = MaskFormerFeatureExtractor.from_pretrained(model_name)
-        print(type(processor))
+        processor = MaskFormerImageProcessor.from_pretrained(model_name)
+        # print(type(processor))
         model = MaskFormerForInstanceSegmentation.from_pretrained(model_name)
 
         inputs = processor(images=image, return_tensors="pt")
 
         outputs = model(**inputs)
-        print(type(outputs))
-        print(outputs)
+        results = processor.post_process_panoptic_segmentation(outputs, target_sizes=[image.size[::-1]])[0]
+        
         pass
     else:
         raise NameError("Model is not implemented")

understand.py

@@ -63,6 +63,67 @@ results = processor.post_process_panoptic_segmentation(outputs, target_sizes=[im
 # type(results["segmentation"]) --> <class 'torch.Tensor'>
 
 
+def show_mask_for_number(map_to_use, label_id):
+
+    if torch.cuda.is_available():
+        mask = (map_to_use.cpu().numpy() == label_id)
+    else:
+        mask = (map_to_use.numpy() == label_id)
+    
+    visual_mask = (mask* 255).astype(np.uint8)
+    visual_mask = Image.fromarray(visual_mask)
+    plt.imshow(visual_mask)
+    plt.show()
+
+
+def get_coordinates_for_bb_simple(map_to_use, label_id):
+    if torch.cuda.is_available():
+        mask = (map_to_use.cpu().numpy() == label_id)
+    else:
+        mask = (map_to_use.numpy() == label_id)
+    
+    x, y = np.where(mask==True)
+    x_max, x_min = max(x), min(x)
+    y_max, y_min = max(y), min(y)
+    return (x_min, y_min), (x_max, y_max)
+
+def make_simple_box(left_top, right_bottom, map_size):
+    full_mask = np.full(map_size, False)
+    left_x, top_y = left_top
+    right_x, bottom_y = right_bottom
+    full_mask[left_x:right_x, top_y] = True
+    full_mask[left_x:right_x, bottom_y] = True
+    full_mask[left_x, top_y:bottom_y] = True
+    full_mask[right_x, top_y:bottom_y] = True
+
+    visual_mask = (full_mask* 255).astype(np.uint8)
+    visual_mask = Image.fromarray(visual_mask)
+    plt.imshow(visual_mask)
+    plt.show()
+
+
+def test(map_to_use, label_id):
+    if torch.cuda.is_available():
+        mask = (map_to_use.cpu().numpy() == label_id)
+    else:
+        mask = (map_to_use.numpy() == label_id)
+    
+    
+    lt, rb = get_coordinates_for_bb_simple(map_to_use, label_id)
+    left_x, top_y = lt
+    right_x, bottom_y = rb
+    
+    mask[left_x:right_x, top_y] = .5
+    mask[left_x:right_x, bottom_y] = .5
+    mask[left_x, top_y:bottom_y] = .5
+    mask[right_x, top_y:bottom_y] = .5
+
+    visual_mask = (mask* 255).astype(np.uint8)
+    visual_mask = Image.fromarray(visual_mask)
+    plt.imshow(visual_mask)
+    plt.show()
+
+
 
 # From Tutorial (Box 79)
 # def get_mask(segment_idx):
@@ -129,4 +190,17 @@ array([[False, False, False, ..., False, False, False],
 >>> results["segments_info"][0]
 {'id': 1, 'label_id': 25, 'was_fused': False, 'score': 0.998022}
 >>> 
+"""
+
+"""
+>>> np.where(mask==True)
+(array([300, 300, 300, ..., 392, 392, 392]), array([452, 453, 454, ..., 473, 474, 475]))
+>>> max(np.where(mask==True)[0])
+392
+>>> min(np.where(mask==True)[0])
+300
+>>> max(np.where(mask==True)[1])
+538
+>>> min(np.where(mask==True)[1])
+399
 """