fix of calc crops

Author: Koldim2001

patched_yolo_infer/README.md

@@ -201,7 +201,7 @@ classes_names=result.filtered_classes_names
 
 ## __How to automatically determine optimal parameters for patches (crops):__
 
-To efficiently process a large number of images of varying sizes and contents, manually selecting the optimal patch sizes and overlaps can be cumbersome. To address this, an algorithm has been developed to automatically calculate the best parameters for patches (crops).
+To efficiently process a large number of images of varying sizes and contents, manually selecting the optimal patch sizes and overlaps can be difficult. To address this, an algorithm has been developed to automatically calculate the best parameters for patches (crops).
 
 The  `auto_calculate_crop_values` function operates in two modes:
 
@@ -221,6 +221,6 @@ img = cv2.imread(img_path)
 
 # Calculate the optimal crop size and overlap for an image
 shape_x, shape_y, overlap_x, overlap_y = auto_calculate_crop_values(
-    image=img, mode="network_analysis", model=YOLO("yolov8m.pt")
+    image=img, mode="network_based", model=YOLO("yolov8m.pt")
 )
 ```

patched_yolo_infer/functions_extra.py

@@ -503,7 +503,7 @@ def basic_crop_size_calculation(width, height):
     return crop_shape_x, crop_shape_y, crop_overlap_x, crop_overlap_y
 
 
-def auto_calculate_crop_values(image, mode="network_analysis", model=None, classes_list=None):
+def auto_calculate_crop_values(image, mode="network_based", model=None, classes_list=None):
     """
     Automatically calculate the optimal crop size and overlap for an image.
 
@@ -513,7 +513,7 @@ def auto_calculate_crop_values(image, mode="network_analysis", model=None, class
 
     Parameters:
     image (numpy.ndarray): The input BGR image.
-    mode (str): The type of analysis to perform. Can be "image_size_analysis" or "network_analysis". 
+    mode (str): The type of analysis to perform. Can be "resolution_based" or "network_based". 
         Default is "network_analysis".
     model (YOLO): The YOLO model to use for object detection. If None, a default model yolov8m
         will be loaded. Default is None.
@@ -527,8 +527,10 @@ def auto_calculate_crop_values(image, mode="network_analysis", model=None, class
     height, width = image.shape[:2]
 
     # If the mode is 'image_size_analysis', calculate crop size based on image dimensions
-    if mode == 'image_size_analysis':
-        crop_shape_x, crop_shape_y, crop_overlap_x, crop_overlap_y = basic_crop_size_calculation(width, height)  
+    if mode == 'resolution_based':
+        crop_shape_x, crop_shape_y, crop_overlap_x, crop_overlap_y = basic_crop_size_calculation(
+            width, height
+        )
     else:
         # If no model is provided, load a default YOLO model
         if model is None:
@@ -556,7 +558,7 @@ def auto_calculate_crop_values(image, mode="network_analysis", model=None, class
                 max_height = box_height
 
         # Determine the maximum dimension (width or height) of the detected objects
-        max_value = max(box_height, max_width)
+        max_value = max(max_width, max_height)
 
         # Adjust crop size and overlap based on the maximum detected object dimension
         if width > height:
@@ -584,4 +586,12 @@ def auto_calculate_crop_values(image, mode="network_analysis", model=None, class
         if width // crop_shape_x > 7:
             crop_shape_x = width // 7
 
+        # Handling cases where patches are not needed along the axes
+        if height / crop_shape_y < 1.25:
+            crop_shape_y = height
+            crop_overlap_y = 0
+        if width / crop_shape_x < 1.25:
+            crop_shape_x = width
+            crop_overlap_x = 0
+
     return crop_shape_x, crop_shape_y, crop_overlap_x, crop_overlap_y