Improve detection label placement

detection_utils.py

@@ -48,6 +48,48 @@ def detection_color(class_id: int) -> tuple[int, int, int]:
     )
 
 
+def _boxes_overlap(
+    first: tuple[float, float, float, float],
+    second: tuple[float, float, float, float],
+) -> bool:
+    return not (
+        first[2] <= second[0]
+        or second[2] <= first[0]
+        or first[3] <= second[1]
+        or second[3] <= first[1]
+    )
+
+
+def _label_box(
+    object_box: list[float],
+    label_width: float,
+    label_height: float,
+    image_size: tuple[int, int],
+    occupied: list[tuple[float, float, float, float]],
+) -> tuple[float, float, float, float]:
+    """Place a label near its object while avoiding earlier labels."""
+    image_width, image_height = image_size
+    center_x = (object_box[0] + object_box[2]) / 2
+    x = max(0.0, min(center_x - label_width / 2, image_width - label_width))
+
+    candidates: list[tuple[float, float, float, float]] = []
+    for slot in range(7):
+        bottom = object_box[1] - 4 - slot * (label_height + 4)
+        top = bottom - label_height
+        if top >= 0:
+            candidates.append((x, top, x + label_width, bottom))
+    for slot in range(4):
+        top = object_box[3] + 4 + slot * (label_height + 4)
+        bottom = top + label_height
+        if bottom <= image_height:
+            candidates.append((x, top, x + label_width, bottom))
+
+    for candidate in candidates:
+        if not any(_boxes_overlap(candidate, used) for used in occupied):
+            return candidate
+    return candidates[0] if candidates else (x, 0.0, x + label_width, label_height)
+
+
 def render_detection(
     image: Image.Image,
     detections: Iterable[dict[str, object]],
@@ -58,8 +100,13 @@ def render_detection(
     short_side = min(rendered.size)
     line_width = max(2, round(short_side / 320))
     font = _load_font(max(13, min(24, round(short_side / 55))))
+    occupied_labels: list[tuple[float, float, float, float]] = []
 
-    for detection in detections:
+    ordered_detections = sorted(
+        detections,
+        key=lambda item: (float(item["x1"]), float(item["y1"])),
+    )
+    for detection in ordered_detections:
         class_id = int(detection["class_id"])
         color = detection_color(class_id)
         box = [
@@ -73,13 +120,24 @@ def render_detection(
             f"{float(detection['confidence']):.2f}"
         )
         draw.rectangle(box, outline=color, width=line_width)
-        text_box = draw.textbbox((box[0], box[1]), label, font=font)
+        text_box = draw.textbbox((0, 0), label, font=font)
         text_height = text_box[3] - text_box[1]
         text_width = text_box[2] - text_box[0]
-        text_y = max(0.0, box[1] - text_height - 8)
-        background = [box[0], text_y, box[0] + text_width + 8, text_y + text_height + 8]
+        background = _label_box(
+            box,
+            text_width + 8,
+            text_height + 8,
+            rendered.size,
+            occupied_labels,
+        )
+        occupied_labels.append(background)
         draw.rectangle(background, fill=color)
-        draw.text((box[0] + 4, text_y + 4), label, fill="white", font=font)
+        draw.text(
+            (background[0] + 4, background[1] + 4),
+            label,
+            fill="white",
+            font=font,
+        )
 
     return rendered
 

tests/test_detection_utils.py

@@ -55,3 +55,9 @@ def test_detection_overlay_matches_input_size():
     rendered = utils.render_detection(image, DETECTIONS)
     assert rendered.size == image.size
     assert rendered.getpixel((1, 2)) != (255, 255, 255)
+
+
+def test_label_boxes_avoid_existing_labels():
+    first = utils._label_box([20, 30, 30, 40], 20, 10, (100, 100), [])
+    second = utils._label_box([22, 30, 32, 40], 20, 10, (100, 100), [first])
+    assert not utils._boxes_overlap(first, second)