vision-agent 0.2.107__tar.gz → 0.2.109__tar.gz

{vision_agent-0.2.107 → vision_agent-0.2.109}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vision-agent
-Version: 0.2.107
+Version: 0.2.109
 Summary: Toolset for Vision Agent
 Author: Landing AI
 Author-email: dev@landing.ai

{vision_agent-0.2.107 → vision_agent-0.2.109}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "vision-agent"
-version = "0.2.107"
+version = "0.2.109"
 description = "Toolset for Vision Agent"
 authors = ["Landing AI <dev@landing.ai>"]
 readme = "README.md"

{vision_agent-0.2.107 → vision_agent-0.2.109}/vision_agent/tools/tools.py

@@ -157,7 +157,7 @@ def owl_v2(
     image_size = image.shape[:2]
     image_b64 = convert_to_b64(image)
     request_data = {
-        "prompts": prompt.split(","),
+        "prompts": [s.strip() for s in prompt.split(",")],
         "image": image_b64,
         "confidence": box_threshold,
         "function_name": "owl_v2",
@@ -312,6 +312,8 @@ def ocr(image: np.ndarray) -> List[Dict[str, Any]]:
 
     pil_image = Image.fromarray(image).convert("RGB")
     image_size = pil_image.size[::-1]
+    if image_size[0] < 1 and image_size[1] < 1:
+        return []
     image_buffer = io.BytesIO()
     pil_image.save(image_buffer, format="PNG")
     buffer_bytes = image_buffer.getvalue()