Notebook for Streaming Service

ai_services/vision/python/stream-video/stream-process-output.ipynb

@@ -0,0 +1,192 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "bacae64c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from oci.object_storage import ObjectStorageClient\n",
+    "import time\n",
+    "import base64\n",
+    "from PIL import Image\n",
+    "from io import BytesIO\n",
+    "import cv2\n",
+    "import numpy as np\n",
+    "import json\n",
+    "import oci\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "019e28f5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "NAMESPACE = \"\"\n",
+    "BUCKET = \"\"\n",
+    "PREFIX = \"\"\n",
+    "CONFIG_PROFILE = \"\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "20ecb251",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Configure OCI client\n",
+    "config = oci.config.from_file('~/.oci/config', profile_name=CONFIG_PROFILE)\n",
+    "endpoint = \"https://objectstorage.us-ashburn-1.oraclecloud.com\"\n",
+    "token_file = config['security_token_file']\n",
+    "with open(token_file, 'r') as f:\n",
+    "    token = f.read()\n",
+    "\n",
+    "private_key = oci.signer.load_private_key_from_file(config['key_file'])\n",
+    "signer = oci.auth.signers.SecurityTokenSigner(token, private_key)\n",
+    "object_storage_client = ObjectStorageClient(config=config, signer=signer, service_endpoint=endpoint)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "ccf81481",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def decode_image(image_data):\n",
+    "    \"\"\"Decode base64 image to OpenCV format and ensure it has 3 channels (RGB).\"\"\"\n",
+    "    image_bytes = base64.b64decode(image_data)\n",
+    "    image = Image.open(BytesIO(image_bytes))\n",
+    "\n",
+    "    if image.mode not in (\"RGB\", \"RGBA\"):\n",
+    "        image = image.convert(\"RGB\")\n",
+    "\n",
+    "    return cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "81e8fb29",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def draw_faces(image, faces, frame):\n",
+    "    \"\"\"Draw bounding boxes around detected faces.\"\"\"\n",
+    "    height, width, _ = image.shape\n",
+    "    for face in faces:\n",
+    "        if face['name'] == 'weapon':\n",
+    "            vertices = face['boundingPolygon']['normalizedVertices']\n",
+    "            pts = [(int(v['x'] * width), int(v['y'] * height)) for v in vertices]\n",
+    "            cv2.polylines(image, [np.array(pts)], isClosed=True, color=(0, 255, 0), thickness=2)\n",
+    "    if frame is not None:\n",
+    "        # Add frame number text in the top right corner\n",
+    "        text = f\"Frame: {frame}\"\n",
+    "        font = cv2.FONT_HERSHEY_SIMPLEX\n",
+    "        font_scale = 1\n",
+    "        color = (0, 255, 0)\n",
+    "        thickness = 1\n",
+    "        text_size = cv2.getTextSize(text, font, font_scale, thickness)[0]\n",
+    "        text_x = width - text_size[0] - 10\n",
+    "        text_y = 30\n",
+    "        cv2.putText(image, text, (text_x, text_y), font, font_scale, color, thickness)\n",
+    "\n",
+    "    return image"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "299aa92f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def process_and_display(message):\n",
+    "    \"\"\"Process detection data and update the window.\"\"\"\n",
+    "    message = message.replace(\"'\", '\"')  # Make it valid JSON\n",
+    "    message_dict = json.loads(message)\n",
+    "\n",
+    "    image = decode_image(message_dict['imageData'])\n",
+    "    if len(message_dict['detectedObjects']) > 0:\n",
+    "        face_id = message_dict['detectedObjects'][0]['objectId']\n",
+    "        print(\"face detected\", face_id)\n",
+    "\n",
+    "    print(message_dict['detectedObjects'])\n",
+    "    # Uncomment if you want to visualize\n",
+    "    annotated_image = draw_faces(image, message_dict['detectedObjects'], message_dict[\"imageData\"])\n",
+    "    cv2.imshow(\"Face Detection\", annotated_image)\n",
+    "    cv2.waitKey(1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "1fafd752",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def main():\n",
+    "    while True:\n",
+    "        ts_ns = int(time.time() * 1_000_000_000) - 3_000_000_000\n",
+    "        filename = f\"new1/frame_{ts_ns}.json\"\n",
+    "\n",
+    "        objects = object_storage_client.list_objects(\n",
+    "            namespace_name=NAMESPACE,\n",
+    "            bucket_name=BUCKET,\n",
+    "            prefix=PREFIX,\n",
+    "            start_after=filename\n",
+    "        ).data.objects\n",
+    "\n",
+    "        if objects:\n",
+    "            obj_name = objects[0].name\n",
+    "            resp = object_storage_client.get_object(\n",
+    "                namespace_name=NAMESPACE,\n",
+    "                bucket_name=BUCKET,\n",
+    "                object_name=obj_name\n",
+    "            )\n",
+    "            content = resp.data.content.decode(\"utf-8\")\n",
+    "            process_and_display(content)\n",
+    "            print(f\"File: {obj_name}\")\n",
+    "            # print(content)\n",
+    "\n",
+    "        time.sleep(1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "27638749",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "if __name__ == \"__main__\":\n",
+    "    main()"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

ai_services/vision/python/stream-video/stream-process-output.py

@@ -0,0 +1,109 @@
+import time
+import base64
+from PIL import Image
+from io import BytesIO
+import cv2
+import numpy as np
+import json
+import oci
+from oci.object_storage import ObjectStorageClient
+
+
+NAMESPACE = ""
+BUCKET = ""
+PREFIX = ""
+CONFIG_PROFILE = ""
+
+
+def decode_image(image_data):
+    """Decode base64 image to OpenCV format and ensure it has 3 channels (RGB)."""
+    image_bytes = base64.b64decode(image_data)
+    image = Image.open(BytesIO(image_bytes))
+
+    if image.mode not in ("RGB", "RGBA"):
+        image = image.convert("RGB")
+
+    return cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)
+
+
+def process_and_display(message):
+    """Process detection data and update the window."""
+    message = message.replace("'", '"')  # Make it valid JSON
+    message_dict = json.loads(message)
+
+    image = decode_image(message_dict['imageData'])
+    if len(message_dict['detectedObjects']) > 0:
+        face_id = message_dict['detectedObjects'][0]['objectId']
+        print("face detected", face_id)
+
+    print(message_dict['detectedObjects'])
+    # Uncomment if you want to visualize
+    annotated_image = draw_faces(image, message_dict['detectedObjects'], message_dict["imageData"])
+    cv2.imshow("Face Detection", annotated_image)
+    cv2.waitKey(1)
+
+
+def draw_faces(image, faces, frame):
+    """Draw bounding boxes around detected faces."""
+    height, width, _ = image.shape
+    for face in faces:
+        if face['name'] == 'weapon':
+            vertices = face['boundingPolygon']['normalizedVertices']
+            pts = [(int(v['x'] * width), int(v['y'] * height)) for v in vertices]
+            cv2.polylines(image, [np.array(pts)], isClosed=True, color=(0, 255, 0), thickness=2)
+    if frame is not None:
+        # Add frame number text in the top right corner
+        text = f"Frame: {frame}"
+        font = cv2.FONT_HERSHEY_SIMPLEX
+        font_scale = 1
+        color = (0, 255, 0)
+        thickness = 1
+        text_size = cv2.getTextSize(text, font, font_scale, thickness)[0]
+        text_x = width - text_size[0] - 10
+        text_y = 30
+        cv2.putText(image, text, (text_x, text_y), font, font_scale, color, thickness)
+
+    return image
+
+
+def main():
+    # Configure OCI client
+    config = oci.config.from_file('~/.oci/config', profile_name=CONFIG_PROFILE)
+    oendpoint = "https://objectstorage.us-ashburn-1.oraclecloud.com"
+
+    token_file = config['security_token_file']
+    with open(token_file, 'r') as f:
+        token = f.read()
+
+    private_key = oci.signer.load_private_key_from_file(config['key_file'])
+    signer = oci.auth.signers.SecurityTokenSigner(token, private_key)
+    object_storage_client = ObjectStorageClient(config=config, signer=signer, service_endpoint=oendpoint)
+
+    while True:
+        ts_ns = int(time.time() * 1_000_000_000) - 3_000_000_000
+        filename = f"new1/frame_{ts_ns}.json"
+
+        objects = object_storage_client.list_objects(
+            namespace_name=NAMESPACE,
+            bucket_name=BUCKET,
+            prefix=PREFIX,
+            start_after=filename
+        ).data.objects
+
+        if objects:
+            obj_name = objects[0].name
+            resp = object_storage_client.get_object(
+                namespace_name=NAMESPACE,
+                bucket_name=BUCKET,
+                object_name=obj_name
+            )
+            content = resp.data.content.decode("utf-8")
+            process_and_display(content)
+            print(f"File: {obj_name}")
+            # print(content)
+
+        time.sleep(1)
+
+
+if __name__ == "__main__":
+    main()