### Install LandingAI SDK and Set Up Credentials

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/video-analytics-notebook/video-analytics.ipynb

Installs the LandingAI SDK if not already present and sets up API credentials and endpoint configuration. Requires a runtime restart after installation.

```python
api_key       = "land_sk_aMemWbpd41yXnQ0tXvZMh59ISgRuKNRKjJEIUHnkiH32NBJAwf" 
endpoint_id   = "067ce386-3958-4d98-be31-9a9af07e361a"
video_file    = "https://drive.google.com/uc?export=download&id=12I5r1siMRzcejuFxI-izRshBtmWRmVxQ"

# Install LandingAI's SDK only if needed to avoid unnecessary restarts
try:
  import landingai
except ImportError:
  import os
  from IPython.display import display, Markdown
  display(Markdown("## Installing modules. You will need to restart runtime."))
  !pip install landingai
  display(Markdown("## Please restart runtime."))
```

--------------------------------

### Run Video Inference Example

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/capture-service/README.md

Execute this command to start the example application. The program captures frames, runs inference, and displays predictions overlaid on the image. Be mindful of the inference rate limits for free trial accounts.

```bash
python examples/capture-service/run.py 
```

--------------------------------

### Install All Dependencies

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/contributing.md

Installs all project dependencies, including optional extras, using Poetry.

```bash
poetry install --all-extras
```

--------------------------------

### Install Dependencies with Poetry

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/examples.md

Install the necessary Python dependencies, including those for examples, using Poetry.

```bash
poetry install --with examples
```

--------------------------------

### Run RTSP Capture Example

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/examples.md

Execute the streaming video capture example using Python.

```bash
python landingai-python/examples/capture-service/run.py
```

--------------------------------

### Run RTSP Capture Example Locally

Source: https://github.com/landing-ai/landingai-python/blob/main/README.md

Instructions for running the RTSP capture example locally using Git, Poetry, and Python.

```bash
git clone https://github.com/landing-ai/landingai-python.git
poetry install --with examples
poetry shell
python landingai-python/examples/capture-service/run.py
```

--------------------------------

### Install Project Requirements

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/zoom-app/README.md

Run this command in your terminal to install all necessary Python packages for the project. Ensure you have Python 3.9 or higher installed.

```bash
pip install -r requirements.txt
```

--------------------------------

### Install Requirements

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/object-tracking/README.md

Install the necessary Python packages for the object tracking application.

```bash
pip install -r examples/apps/object-tracking/requirements.txt
```

--------------------------------

### Install LandingAI SDK

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/webcam-collab-notebook/webcam-collab-notebook.ipynb

Installs the LandingAI Python SDK if it's not already present. This code block includes a mechanism to force a runtime restart if the installation is successful, which is necessary for the new package to be recognized.

```python
# Install dependencies only if needed to avoid unnecessary restarts
try:
  import landingai
except ImportError:
  import os
  from IPython.display import display, Markdown
  display(Markdown("##Installing modules. You will need to restart runtime."))
  %pip install landingai
  display(Markdown("##Stopping RUNTIME! Please restart runtime."))
  os.kill(os.getpid(), 9) # This is a way to force a restart
```

--------------------------------

### Install Poetry

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/contributing.md

Installs Poetry, a dependency management tool, using a script. Specify a Python version if needed.

```bash
curl -sSL https://install.python-poetry.org | python3 -
```

```bash
curl -sSL https://install.python-poetry.org | python3.10 -
```

--------------------------------

### Complete Pipeline Example: Webcam Monitoring

Source: https://context7.com/landing-ai/landingai-python/llms.txt

An end-to-end example demonstrating webcam monitoring with defect detection. It acquires frames, runs inference, overlays predictions, and saves images when defects are found.

```python
from landingai.pipeline.image_source import Webcam, VideoFile
from landingai.pipeline.frameset import Frame, FrameSet
from landingai.predict import Predictor
import os

# Set API key via environment variable
os.environ["LANDINGAI_API_KEY"] = "land_sk_your_api_key"

# Initialize predictor
predictor = Predictor(endpoint_id="your-endpoint-id")

# Example 1: Webcam monitoring with defect detection
with Webcam(fps=1) as webcam:
    for frame in webcam:
        frame.resize(width=640)
        frame.run_predict(predictor=predictor)

        if "defect" in frame.predictions:
            frame.overlay_predictions()
            frame.save_image("/tmp/defect_detected.png", include_predictions=True)
            print(f"Defect detected! {len(frame.predictions)} issues found")

        # Break after 10 frames for this example
        break
```

--------------------------------

### Install landingai and Download Video

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/license-plate-ocr-notebook/license_plate_ocr.ipynb

Installs the necessary landingai python package and downloads a sample video file for processing. Ensure you have the correct file path for the downloaded video.

```python
!pip install landingai gdown

!gdown "https://drive.google.com/uc?id=16iwE7mcz9zHqKCw2ilx0QEwSCjDdXEW4" -O /tmp/license-plates.mov
```

--------------------------------

### Install Landing AI Python Library

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/post-processings/farmland-coverage/farmland-coverage.ipynb

This section shows the output of a successful installation of the landingai Python package. It also provides a notice about a new pip release and how to upgrade.

```text
Requirement already satisfied: pyOpenSSL<24.0.0,>=16.2.0 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from snowflake-connector-python==3.0.*->landingai==0.0.40) (23.1.1)\nRequirement already satisfied: asn1crypto<2.0.0,>0.24.0 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from snowflake-connector-python==3.0.*->landingai==0.0.40) (1.5.1)\nRequirement already satisfied: idna<4,>=2.5 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from snowflake-connector-python==3.0.*->landingai==0.0.40) (3.4)\nRequirement already satisfied: cffi<2.0.0,>=1.9 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from snowflake-connector-python==3.0.*->landingai==0.0.40) (1.15.1)\nRequirement already satisfied: matplotlib>=3.4.2 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from segmentation-mask-overlay<0.4.0,>=0.3.4->landingai==0.0.40) (3.7.1)\nRequirement already satisfied: python-dotenv>=0.10.4 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from pydantic[dotenv]->landingai==0.0.40) (1.0.0)\nRequirement already satisfied: pycparser in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from cffi<2.0.0,>=1.9->snowflake-connector-python==3.0.*->landingai==0.0.40) (2.21)\nRequirement already satisfied: fonttools>=4.22.0 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from matplotlib>=3.4.2->segmentation-mask-overlay<0.4.0,>=0.3.4->landingai==0.0.40) (4.39.4)\nRequirement already satisfied: pyparsing>=2.3.1 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from matplotlib>=3.4.2->segmentation-mask-overlay<0.4.0,>=0.3.4->landingai==0.0.40) (3.0.9)\nRequirement already satisfied: python-dateutil>=2.7 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from matplotlib>=3.4.2->segmentation-mask-overlay<0.4.0,>=0.3.4->landingai==0.0.40) (2.8.2)\nRequirement already satisfied: cycler>=0.10 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from matplotlib>=3.4.2->segmentation-mask-overlay<0.4.0,>=0.3.4->landingai==0.0.40) (0.11.0)\nRequirement already satisfied: kiwisolver>=1.0.1 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from matplotlib>=3.4.2->segmentation-mask-overlay<0.4.0,>=0.3.4->landingai==0.0.40) (1.4.4)\nRequirement already satisfied: contourpy>=1.0.1 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from matplotlib>=3.4.2->segmentation-mask-overlay<0.4.0,>=0.3.4->landingai==0.0.40) (1.0.7)\nRequirement already satisfied: six>=1.5 in /Users/asia/Library/Caches/pypoetry/virtualenvs/landingai-_c3uY8Dq-py3.10/lib/python3.10/site-packages (from python-dateutil>=2.7->matplotlib>=3.4.2->segmentation-mask-overlay<0.4.0,>=0.3.4->landingai==0.0.40) (1.16.0)\nBuilding wheels for collected packages: landingai\n  Building editable for landingai (pyproject.toml) ... [?25ldone\n[?25h  Created wheel for landingai: filename=landingai-0.0.40-py3-none-any.whl size=4424 sha256=e5c9f30990144126234944e03bc8ddc721f91bade3f58f517a50b570205a2ba4\n  Stored in directory: /private/var/folders/cv/kdkz6y_55z1blffx_f42w6zh0000gn/T/pip-ephem-wheel-cache-qw6dcxys/wheels/57/df/6e/917cfef370ef8a31e12c5b0a81541f8aa462c4f83c9ade0780\nSuccessfully built landingai\nInstalling collected packages: landingai\n  Attempting uninstall: landingai\n    Found existing installation: landingai 0.0.40\n    Uninstalling landingai-0.0.40:\n      Successfully uninstalled landingai-0.0.40\nSuccessfully installed landingai-0.0.40\n\n[notice] A new release of pip available: 22.3.1 -> 23.1.2\n[notice] To update, run: pip install --upgrade pip
```

--------------------------------

### Install LandingAI Python Library

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/index.md

Install the LandingAI Python library using pip. Ensure you are using a compatible version.

```bash
pip install landingai~=0.3.0
```

--------------------------------

### Run Streamlit App

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/object-tracking/README.md

Launch the Streamlit application to start the object detection on a live stream.

```bash
streamlit run examples/apps/object-tracking/main.py
```

--------------------------------

### Install Dependencies (Windows)

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/edge-models/README.md

Installs project dependencies on Windows systems. Note that tflite-runtime is not available for Windows; Tensorflow is recommended.

```bash
pip install -r requirements_win.txt
```

--------------------------------

### Install Dependencies for Image Folder App

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/image-folder-support/README.md

Use this command to install the necessary Python packages for the image folder support application.

```bash
pip install -r examples/apps/image-folder-support/requirements.txt
```

--------------------------------

### Install Requirements

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/surfer-count/README.md

Install the necessary Python packages for the Streamlit app. Ensure you have a requirements.txt file in the specified directory.

```bash
pip install -r examples/apps/surfer-count/requirements.txt
```

--------------------------------

### Install OCR App Dependencies

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/ocr/README.md

Install the necessary Python packages for the OCR app by running this command in your terminal.

```bash
pip install -r examples/apps/ocr/requirements.txt
```

--------------------------------

### Launch Streamlit App for Image Folder Inference

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/image-folder-support/README.md

Run this command to start the Streamlit application for local image folder inference.

```bash
streamlit run examples/apps/image-folder-support/app.py
```

--------------------------------

### Initialize Object Detection Predictor

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/webcam-collab-notebook/webcam-collab-notebook.ipynb

Sets up the Predictor object with a model endpoint and API key. These variables should be configured according to your LandingLens model setup.

```python
#@title Set the following variables as needed for your setup
# These keys correspond to a public LandingLens model
api_key         = "land_sk_aMemWbpd41yXnQ0tXvZMh59ISgRuKNRKjJEIUHnkiH32NBJAwf"  #@param {type:"string"}
model_endpoint  =  "0b3ad59c-80ca-481b-977f-f2470cbe83e3" #@param {type:"string"}
```

```python
from landingai.predict import Predictor
from landingai.visualize import overlay_predictions
        
predictor = Predictor(model_endpoint, api_key=api_key)

# Run prediction and show raw results
results = predictor.predict(frame)
display(results)
```

--------------------------------

### Configure Stream URL and Credentials

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/capture-service/README.md

Update these variables in the `run.py` file with your specific API key, endpoint ID, and RTSP stream URL to customize the example for your camera and deployed model.

```python
api_key = "YOUR_API_KEY"
endpoint_id = "YOUR_ENDPOINT_ID"
stream_url = "YOUR_RTSP_STREAM_URL"
```

--------------------------------

### Install LandingAI with Snowflake Optionals

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/inferences/snowflake-native-app.md

Install the landingai package with the snowflake optionals to enable the SnowflakeNativeAppPredictor class. This is required for running inference with Snowflake Native Apps.

```sh
pip install "landingai[snowflake]"
```

--------------------------------

### Run Streamlit App

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/surfer-count/README.md

Execute the Streamlit application to start the live object detection. This command launches the app in your web browser.

```bash
streamlit run examples/apps/surfer-count/main.py
```

--------------------------------

### Run Edge Model with Hardware Acceleration

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/edge-models/README.md

Executes an edge model bundle with hardware acceleration enabled. The delegate library path is an example and should be replaced with your actual library path.

```bash
python run_bundle_cls.py published_model_bundle.zip image.png /usr/lib/libvx_delegate.so
```

--------------------------------

### Run Inference with Predictor

Source: https://context7.com/landing-ai/landingai-python/llms.txt

Load an image into a Frame object and run prediction using the initialized predictor. This example shows the basic inference process and how to access the number of predictions.

```python
frame = Frame.from_image("/path/to/image.jpg")
frame.run_predict(predictor=predictor)
print(f"Predictions: {len(frame.predictions)}")
```

--------------------------------

### Configure API Key and Endpoint ID

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/post-processings/farmland-coverage/farmland-coverage.ipynb

Set the api_key and endpoint_id variables to access your deployed Visual Prompting model in LandingLens. Ensure these are correctly configured for your setup.

```python
#@title Set the following variables as needed for your setup\napi_key = "land_sk_aMemWbpd41yXnQ0tXvZMh59ISgRuKNRKjJEIUHnkiH32NBJAwf"\nendpoint_id = "63035608-9d24-4342-8042-e4b08e084fde"
```

--------------------------------

### OCR Inference Output Example

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/inferences/ocr.md

This is an example of the expected output format when running OCR inference, showing the extracted text and its associated confidence score.

```text
Hi From (Confidence: 0.9179285764694214)
Landing.AI (Confidence: 0.7755413055419922)
```

--------------------------------

### Use SnowflakeNativeAppPredictor for Inference

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/inferences/snowflake-native-app.md

Example of initializing and using the SnowflakeNativeAppPredictor for inference. Ensure you have the correct endpoint ID, native app URL, and Snowflake credentials. Authentication via private key is recommended for production.

```python
from landingai.pipeline.image_source import Webcam
from landingai.predict import SnowflakeNativeAppPredictor

endpoint_id = "c4344971-fc3c-4cb8-8fd5-5144d25cbd74"
url = "https://focq4dkf-rkwerpo-your-account.snowflakecomputing.app"

predictor = SnowflakeNativeAppPredictor(
    endpoint_id=endpoint_id, # (1)!
    native_app_url=url, # (2)!
    snowflake_account="your-snowflake-account-locator", # (3)!
    snowflake_user="your-snowflake-user",  # (4)!
    snowflake_private_key="your-snowflake-user-private-key",  # (5)!
)

frame = Frame.from_image("/home/user/dataset/some-class/image-1.png")
frame.run_predict(predictor=predictor)
print(frame.predictions)
# [
#   ClassificationPrediction(
#       score=0.9957893490791321,
#       label_name='some-class',
#       label_index=0
#   )
# ]
```

--------------------------------

### Initialize Media Client and List Media

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/metadata.md

Initialize the `Media` client with your project ID and API key. Use the `ls()` method to list all media with metadata from a project. Pagination can be controlled using `offset` and `limit` parameters.

```python
from landingai.data_management import Media

YOUR_API_KEY = "land_sk_12345"
YOUR_PROJECT_ID = 1190

media_client = Media(YOUR_PROJECT_ID, YOUR_API_KEY)
media_client.ls()
```

```python
media_client.ls(offset=0, limit=100)
```

```python
media_client.ls(split="train")
```

--------------------------------

### Serve Documentation Locally

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/_README.md

Run this command to serve the documentation locally. Open http://127.0.0.1:8000/ in your browser to view it.

```bash
poetry run mkdocs serve -f mkdocs.yml
```

--------------------------------

### Initialize Media Client

Source: https://github.com/landing-ai/landingai-python/blob/main/pdocs/user_guide/5_data_management.md

Initialize the Media client with your project ID and API key to interact with LandingLens media management features.

```python
from landingai.data_management import Media

# Provide API Key and project ID
YOUR_API_KEY = "land_sk_12345"
YOUR_PROJECT_ID = 1190

# Lists all medias with metadata from a project
media_client = Media(YOUR_PROJECT_ID, YOUR_API_KEY)
```

--------------------------------

### List Images in Project

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/zoom-app/Zoom_Demo.ipynb

Initialize the Media client for a specific project and list images using offset and limit parameters for pagination.

```python
from landingai.data_management import Media

project_id = 45412657190923

# List images
media_client = Media(project_id, api_key)
media_client.ls(offset=0, limit=2)
```

--------------------------------

### Initialize Predictor and Run Inference on Webcam Stream

Source: https://github.com/landing-ai/landingai-python/blob/main/pdocs/user_guide/4_pipelines.md

Initializes a `Predictor` with your endpoint ID and API key. It then captures frames from a webcam at 1 FPS, resizes them, runs predictions, and saves the resulting image with an overlay of the predictions.

```python
predictor = Predictor(endpoint_id="abcde-1234-xxxx", api_key="land_sk_xxxx")

# Get images from Webcam at a 1 FPS rate, run predictions on it and save results.
with Webcam(fps=1) as webcam:
    for frame in webcam:
        frame
            .downsize(width=512)
            .run_predict(predictor=predictor)
            .save_image(f"/tmp/detected-object", image_src="overlay")
```

--------------------------------

### Install Requirements for Crack Measurer

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/crack-measurer/README.md

Installs the necessary Python packages for the crack measurer application. Ensure you are in the project's root directory.

```bash
pip install -r examples/apps/crack-measurer/requirements.txt
```

--------------------------------

### Load Images from URIs

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/post-processings/farmland-coverage/farmland-coverage.ipynb

Fetches image data from provided URIs and opens them using Pillow. Ensure Pillow is installed (`pip install Pillow`).

```python
from PIL import Image as PILImage
from landingai.storage.data_access import fetch_from_uri

img_paths = [
    fetch_from_uri("https://drive.google.com/uc?id=1z7KiNcXmfbNQL-rz6NMv7eVTBh-CnpAs"),
    fetch_from_uri("https://drive.google.com/uc?id=1Vfbrdj8UFI03dv9fncw2OWWaadGeWBDq"),
    fetch_from_uri("https://drive.google.com/uc?id=1kauz5WyB3rJb0_9qM1MukVmHkNJ9t1vQ"),
]

images = [PILImage.open(path) for path in img_paths]
```

--------------------------------

### Install Requirements for License Plate OCR App

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/license-plate-ocr/README.md

Install the necessary Python packages for the license plate OCR Streamlit app by running this command.

```bash
pip install -r examples/apps/license-plate-ocr/requirements.txt
```

--------------------------------

### Run Edge Model (Basic)

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/edge-models/README.md

Executes an edge model bundle with a specified image. This is the basic execution command.

```bash
python run_bundle_cls.py published_model_bundle.zip image.png
```

--------------------------------

### Run the Streamlit Application

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/zoom-app/README.md

Execute this command to launch the Landing AI application using Streamlit. The application will open in your web browser. You may need to grant camera permissions.

```bash
streamlit run app.py
```

--------------------------------

### Set Python Version with Poetry

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/contributing.md

Activates a specific Python version for your Poetry environment after Poetry has been installed.

```bash
poetry env use 3.10
```

--------------------------------

### List Media with Default Pagination

Source: https://github.com/landing-ai/landingai-python/blob/main/pdocs/user_guide/5_data_management.md

List all media assets in a project with default pagination settings. The output includes metadata for each media item.

```python
media_client.ls()
# Output:
# { medias: [{'id': 4358352, 'mediaType': 'image', 'srcType': 'user', 'srcName': 'Michal', 'properties': {'width': 258, 'height': 176}, 'name': 'n01443537_501.JPEG', 'uploadTime': '2020-09-15T22:29:01.338Z', 'metadata': {'split': 'train', 'source': 'prod'}, 'media_status': 'raw'}, ...],
# num_requested: 1000,
# count: 300,
# offset: 0,
# limit: 1000 }
```

--------------------------------

### Capture and Resize Desktop Screenshots to Video

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/image-acquisition/screenshots.md

Capture the first 20 frames from your desktop, resize each frame to a width of 512 pixels while maintaining aspect ratio, and save the processed frames as a video file. Ensure you have the necessary imports and initialize `FrameSet` before capturing.

```python
from landingai.pipeline.image_source import Screenshot
from landingai.pipeline.frameset import FrameSet
import time

frameset = FrameSet() # (1)!
with Screenshot() as screenshots: # (2)!
    for i, frame in enumerate(screenshots): # (3)!
        if i >= 20:
            break
        frame.resize(width=512) # (4)!
        frameset.append(frame) # (5)!
        time.sleep(0.5) # (6)!
frameset.save_video("/tmp/resized-video.mp4") # (7)!
```

--------------------------------

### Import Core SDK Components

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/zoom-app/Zoom_Demo.ipynb

Import necessary modules from the LandingAI Python SDK for camera, prediction, and image operations.

```python
from landingai.pipeline.image_source import NetworkedCamera
from landingai.predict import Predictor
from landingai.image_source_ops import take_photo_from_webcam
from landingai.visualize import overlay_bboxes
```

--------------------------------

### Acquire Image from File

Source: https://github.com/landing-ai/landingai-python/blob/main/README.md

Load an image from a file path, resize it, and save the resized image. Ensure the image file exists at the specified path.

```python
from landingai.pipeline.frameset import Frame

frame = Frame.from_image("/path/to/your/image.jpg")
frame.resize(width=512, height=512)
frame.save_image("/tmp/resized-image.png")
```

--------------------------------

### Run Inference on Single Image with FrameSet

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/zoom-app/Zoom_Demo.ipynb

Process a single image file using the FrameSet class for inference. This example demonstrates chaining operations for prediction, visualization, and display.

```python
from landingai.pipeline.frameset import FrameSet

# Run inference on a single image
frs = FrameSet.from_image("image.jpg")
frs.run_predict(predictor=model).overlay_predictions().show_image().show_image(image_src="overlay")
```

--------------------------------

### Run Inference on Webcam Frames with LandingAI SDK

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/inferences/frames-inference.md

This snippet demonstrates how to capture frames from a webcam, resize them, and run predictions using the LandingAI SDK. It checks for a specific label ('coffee-mug') in the predictions and prints them if found. Ensure you have your API key and endpoint ID.

```python
from landingai.pipeline.image_source import Webcam
from landingai.predict import Predictor

predictor = Predictor(
    endpoint_id="<insert your endpoint ID here>",
    api_key="<insert your API key here>",
)
with Webcam(fps=0.5) as webcam:
    for frame in webcam:
        frame.resize(width=512)
        frame.run_predict(predictor=predictor)
        if "coffee-mug" in frame.predictions:
            print(frame.predictions)
```

--------------------------------

### Iterate, Resize, and Save Video Frames

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/image-acquisition/video-file.md

Use `landingai.image_source.VideoFile` to iterate over video frames. Resize each frame and collect them in a `FrameSet` before saving to a new video file. Specify `samples_per_second` when opening the video file.

```python
from landingai.pipeline.image_source import VideoFile
from landingai.pipeline.frameset import FrameSet

frameset = FrameSet() # (1)!
with VideoFile("/path/to/your/file.mp4", samples_per_second=5) as video_file: # (2)!
    for frame in video_file: # (3)!
        frame.resize(width=256) # (4)!
        frameset.append(frame) # (5)!
frameset.save_video("/tmp/resized-video.mp4") # (6)!
```

--------------------------------

### Get Label Map for Segmentation Masks

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/metadata.md

Use the `Label` API to retrieve the label map required for uploading segmentation masks. This map associates pixel values with class names and should be saved to a JSON file.

```python
client = Label(project_id, api_key)
client.get_label_map()
{'0': 'ok', '1': 'cat', '2': 'dog'}  # then write this map to a JSON file locally
```

--------------------------------

### Clone Repository with Git

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/examples.md

Use this command to clone the LandingAI Python SDK repository to your local machine.

```bash
git clone https://github.com/landing-ai/landingai-python.git
```

--------------------------------

### Filter Predictions by Confidence Score using LandingAI SDK

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/inferences/frames-inference.md

This example demonstrates how to filter predictions to include only those with a confidence score greater than 0.95 for the 'coffee-mug' label. It prints a message if such a prediction is found, indicating high confidence.

```python
with Webcam(fps=0.5) as webcam:
    for frame in webcam:
        frame.run_predict(predictor=predictor)
        if "coffee-mug" in frame.predictions.filter_threshold(0.95):
            print(f"I'm super sure I found a coffee-mug in the frame!")
```

--------------------------------

### Run Inference with EdgePredictor

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/inferences/docker-deployment.md

Use `EdgePredictor` to send images to a Docker container for inference. Specify the host and port of your running container. This example captures frames from a webcam, runs predictions, overlays them, and saves an image if a 'coffee-mug' is detected.

```python
from landingai.pipeline.image_source import Webcam
from landingai.predict import EdgePredictor

predictor = EdgePredictor(host="localhost", port=8000) # (1)!
with Webcam(fps=15) as webcam: # (2)!
    for frame in webcam:
        frame.run_predict(predictor=predictor) # (3)!
        frame.overlay_predictions()
        if "coffee-mug" in frame.predictions:  # (4)!
            frame.save_image(
                "/tmp/latest-webcam-image.png",
                include_predictions=True
            )
```

--------------------------------

### List and Query Media in LandingLens

Source: https://context7.com/landing-ai/landingai-python/llms.txt

Provides functionality to list and filter media within a LandingLens project. Supports pagination, filtering by status, metadata, and updating split assignments.

```python
from landingai.data_management.media import Media

media_client = Media(project_id=12345, api_key="land_sk_your_api_key")

# List all medias (with pagination)
result = media_client.ls(offset=0, limit=100)
print(f"Found {result['count']} medias")

for media in result['medias']:
    print(f"ID: {media['id']}, Name: {media['name']}")

# Filter by media status
result = media_client.ls(media_status="approved")
result = media_client.ls(media_status=["approved", "raw"])

# Filter by metadata
result = media_client.ls(batch_id="batch_001", camera_id="camera_A")

# Update split assignments
media_ids = [1001, 1002, 1003]
media_client.update_split_key(media_ids, split_key="test")

# Remove from split (set to unassigned)
media_client.update_split_key(media_ids, split_key="")
```

--------------------------------

### Display an Image

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/post-processings/farmland-coverage/farmland-coverage.ipynb

Displays the first image loaded from the specified paths. This is useful for a quick visual check.

```python
display(images[0])
```

--------------------------------

### Initialize SnowflakeNativeAppPredictor with Private Key

Source: https://context7.com/landing-ai/landingai-python/llms.txt

Initialize the SnowflakeNativeAppPredictor using Snowflake account credentials and a private key for authentication. This is an alternative to password authentication. Ensure the private key is correctly formatted.

```python
predictor = SnowflakeNativeAppPredictor(
    endpoint_id="your-endpoint-id",
    snowflake_account="your_account.region",
    snowflake_user="your_username",
    snowflake_private_key="-----BEGIN PRIVATE KEY-----\n...\n-----END PRIVATE KEY-----",
    native_app_url="https://your-native-app-url.snowflakecomputing.app"
)
```

--------------------------------

### List Media Information

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/metadata.md

Use the media_client.ls() function to retrieve a list of media items within a project, including their IDs, metadata, and upload times.

```python
media_client.ls()
>>> {
 medias: [
  {
   'id': 4358352,
   'mediaType': 'image',
   'srcType': 'user',
   'srcName': 'Michal',
   'properties': {'width': 258 'height': 176},
   'name': 'n01443537_501.JPEG',
   'uploadTime': '2020-09-15T22:29:01.338Z',
   'metadata': {'split': 'train' 'source': 'prod'},
   'media_status': 'raw'
  },
  ...
 ],
 num_requested: 1000,
 count: 300,
 offset: 0,
 limit: 1000
}
```

--------------------------------

### Overlay Predictions on Webcam Feed

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/inferences/overlaying-predictions.md

Initializes a Predictor, iterates through webcam frames, runs inference, overlays predictions, and saves the frame if a specific object is detected. Ensure your endpoint ID and API key are correctly set.

```python
from landingai.pipeline.image_source import Webcam
from landingai.predict import Predictor

predictor = Predictor( # (1)!
    endpoint_id="<insert your endpoint ID here>",
    api_key="<insert your API key here>",
)

with Webcam(fps=0.5) as webcam:
    for frame in webcam: # (2)!
        frame.resize(width=512)
        frame.run_predict(predictor=predictor) # (3)!
        frame.overlay_predictions() # (4)!
        if "coffee-mug" in frame.predictions:
            frame.save_image(
                path="frame-with-overlay.png",
                include_predictions=True, # (5)!
            )

```

--------------------------------

### Upload Folder of Images

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/metadata.md

Upload an entire folder of images to your project. Assign all uploaded images to a specified split, such as 'train'.

```python
media_client.upload("/Users/tom/Downloads/images", split="train")
```

--------------------------------

### Run Inference on Webcam Frames

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/index.md

Capture frames from a webcam, send them for inference to a deployed LandingLens model, and save images with predictions if a specific object is detected. Resizing frames before inference can save bandwidth.

```python
from landingai.pipeline.image_source import Webcam
from landingai.predict import Predictor

predictor = Predictor(  # (1)!
    endpoint_id="abcdef01-abcd-abcd-abcd-01234567890", # (2)!
    api_key="land_sk_xxxxxx", # (3)!
)
with Webcam(fps=0.5) as webcam:
    for frame in webcam:
        frame.resize(width=512) # (4)!
        frame.run_predict(predictor=predictor) # (5)!
        frame.overlay_predictions()
        if "coffee-mug" in frame.predictions:  # (6)!
            frame.save_image("/tmp/latest-webcam-image.png", include_predictions=True) # (7)!
```

--------------------------------

### Create Snowflake User with Private Key Authentication

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/inferences/snowflake-native-app.md

SQL commands to create a Snowflake user and role for LandingAI library integration, using RSA public key for authentication. This is the preferred method for production environments.

```sql
CREATE OR REPLACE USER LANDING_LIBRARY_USER
    LOGIN_NAME = 'LANDING_LIBRARY_USER'
    DISPLAY_NAME = 'LandingAI Library User'
    COMMENT = 'User for LandingAI library'
    RSA_PUBLIC_KEY = 'MIIBI...';  -- Replace this with your public key
CREATE ROLE LANDINGLENS_EXTERNAL_ACCESS;
GRANT ROLE LANDINGLENS_EXTERNAL_ACCESS
    TO USER LANDING_LIBRARY_USER;
GRANT APPLICATION ROLE llens_snw_production.LLENS_PUBLIC
    TO ROLE LANDINGLENS_EXTERNAL_ACCESS;
```

--------------------------------

### Assign Metadata Values to Images

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/metadata.md

Use the Metadata client to assign values to predefined metadata keys for specific images. Ensure metadata keys are created in the LandingLens UI beforehand.

```python
from landingai.data_management import Metadata

# Provide the API key and project ID
YOUR_API_KEY = "land_sk_12345"
YOUR_PROJECT_ID = 1190
metadata_client = Metadata(YOUR_PROJECT_ID, YOUR_API_KEY)
# Set three metadata values ('timestamp', 'country' and 'labeler') for images with IDs 123 and 124
metadata_client.update(media_ids=[123, 124], timestamp=12345, country="us", labeler="tom")
# Output:
# {
#    project_id: 1190,
#    metadata:   {"country": "us", "timestamp": "12345", labeler="tom"},
#    media_ids:  [123, 124]
# }
```

--------------------------------

### Acquire Images from Webcam

Source: https://github.com/landing-ai/landingai-python/blob/main/README.md

Continuously capture frames from a webcam, resize them, and save each frame as an image. Adjust the FPS for desired capture rate.

```python
from landingai.pipeline.image_source import Webcam

with Webcam(fps=0.5) as webcam:
    for frame in webcam:
        frame.resize(width=512, height=512)
        frame.save_image("/tmp/webcam-image.png")
```

--------------------------------

### Predict with Webcam Image

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/zoom-app/Zoom_Demo.ipynb

Capture an image from your webcam and perform inference using the initialized predictor. The results are overlaid on the image.

```python
# Grab a frame frome your webcam and run predict on it
image = take_photo_from_webcam()
prediction = model.predict(image)
overlay_bboxes(prediction, image)
```

--------------------------------

### Run Inference on Webcam Feed

Source: https://github.com/landing-ai/landingai-python/blob/main/README.md

Capture frames from a webcam, resize them, and send them for inference to a deployed LandingLens model. Requires an endpoint ID and API key. Predictions are overlaid and saved if a specific object is detected.

```python
from landingai.pipeline.image_source import Webcam
from landingai.predict import Predictor

predictor = Predictor(
    endpoint_id="abcdef01-abcd-abcd-abcd-01234567890",
    api_key="land_sk_xxxxxx",
)
with Webcam(fps=0.5) as webcam:
    for frame in webcam:
        frame.resize(width=512)
        frame.run_predict(predictor=predictor)
        frame.overlay_predictions()
        if "coffee-mug" in frame.predictions:
            frame.save_image("/tmp/latest-webcam-image.png", include_predictions=True)
```

--------------------------------

### Upload Folder of Images

Source: https://github.com/landing-ai/landingai-python/blob/main/pdocs/user_guide/5_data_management.md

Upload an entire folder of images to LandingLens. All images within the folder will be uploaded and assigned to the specified split.

```python
# upload a folder of images and assign them to the 'train' split
media_client.upload("/Users/tom/Downloads/images", split="train")
```

--------------------------------

### Activate Poetry Shell

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/examples.md

Activate the virtual environment created by Poetry to manage project dependencies.

```bash
poetry shell
```

--------------------------------

### Upload Image to Project

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/zoom-app/Zoom_Demo.ipynb

Upload a local image file to your LandingAI project, specifying the desired split (e.g., 'dev'). This requires the asyncio loop to be applied, especially in Jupyter environments.

```python
# This code is needed to run asyncio loops in jupyter notebooks, it is not needed if
# you are running python from the terminal.
import nest_asyncio
nest_asyncio.apply()

# Upload image, make sure you have an image in your local directory called image.jpg
media_client.upload("image.jpg", split="dev")
```

--------------------------------

### List Media Items

Source: https://github.com/landing-ai/landingai-python/blob/main/pdocs/user_guide/5_data_management.md

Retrieve a list of media items within a project using the media_client.ls() function. This can be used to find media IDs for other operations.

```python
media_client.ls()
>>> { medias: [{'id': 4358352, 'mediaType': 'image', 'srcType': 'user', 'srcName': 'Michal', 'properties': {'width': 258 'height': 176}, 'name': 'n01443537_501.JPEG', 'uploadTime': '2020-09-15T22:29:01.338Z', 'metadata': {'split': 'train' 'source': 'prod'}, 'media_status': 'raw'}, ...], num_requested: 1000, count: 300, offset: 0, limit: 1000 }
```

--------------------------------

### Create and Use LandingAI FrameSet

Source: https://context7.com/landing-ai/landingai-python/llms.txt

Create a FrameSet, a collection for multiple frames, from an image or numpy array. Supports iteration and accessing individual frames.

```python
from landingai.pipeline.frameset import Frame, FrameSet

# Create a FrameSet from an image
frameset = FrameSet.from_image("/path/to/image.jpg")
print(f"Number of frames: {len(frameset)}")

# Create FrameSet from numpy array
frameset = FrameSet.from_array(numpy_array, is_bgr=True)

# Iterate over frames
for frame in frameset:
    print(frame.image.size)

# Access individual frames
first_frame = frameset[0]
```

--------------------------------

### List Media with Custom Pagination

Source: https://github.com/landing-ai/landingai-python/blob/main/pdocs/user_guide/5_data_management.md

Paginate through media assets using custom offset and limit values to control the number of items retrieved per request.

```python
# paginate with a custom page size
media_client.ls(offset=0, limit=100)
```

--------------------------------

### Launch OCR App

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/ocr/README.md

Run the Streamlit OCR application from your terminal using this command.

```bash
streamlit run examples/apps/ocr/app.py
```

--------------------------------

### Run Inference with Predictor

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/post-processings/farmland-coverage/farmland-coverage.ipynb

Initializes a Predictor with an endpoint ID and API key, then runs predictions on a list of images. The results are aggregated into a single list.

```python
from landingai.predict import Predictor

predictor = Predictor(endpoint_id, api_key=api_key)
predictions = []
for img in images:
        predictions.extend(predictor.predict(img))
```

--------------------------------

### Run Crack Measurer Streamlit App

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/crack-measurer/README.md

Launches the Streamlit application for the crack measurer. This command should be run from the project's root directory.

```bash
streamlit run examples/apps/crack-measurer/app.py
```

--------------------------------

### Set API Key and Model Endpoint

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/rtsp-capture-notebook/rtsp-capture.ipynb

Configure your LandingLens API key and model endpoint. Ensure these are correctly set for authentication and model access.

```python
#@title Set the following variables as needed for your setup
api_key         = "land_sk_aMemWbpd41yXnQ0tXvZMh59ISgRuKNRKjJEIUHnkiH32NBJAwf"  #@param {type:"string"}
model_endpoint  =  "881e04f4-1a16-4728-b19c-a6900d0b8d98" #@param {type:"string"}
```

--------------------------------

### Acquire Image from URI

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/rtsp-capture-notebook/rtsp-capture.ipynb

Fetches an image from a given URI. This can be a direct image file URL or a Google Drive link. For real-time video, set camera_url to 0 for webcam or a specific RTSP stream URL.

```python
from landingai.storage.data_access import fetch_from_uri

# Below are some sample images for you to try out.
img_paths = [
    str(fetch_from_uri("https://drive.google.com/uc?id=1hA1npgptlvdRvLGewMNRqPjfclHHOiZ9")),
    str(fetch_from_uri("https://drive.google.com/uc?id=1vv48yWe0_SwcgX8AgfYg9lxqQ8aNxz0I")),
    str(fetch_from_uri("https://drive.google.com/uc?id=14RKAe-PioJwMf6DZP0Nk03YESJ0h7jb3")),
    str(fetch_from_uri("https://drive.google.com/uc?id=1Q1Y6oM2_63gj35z6h8Qu3j7CVr-bfLtt")),
]

# In order to find the URL for your camera, this is a good start https://www.ispyconnect.com/cameras
# camera_url      =  "rtsp://172.25.101.151/ch0_0.h264" #@param {type:"string"}
camera_url      =  img_paths[0] # Use sample images
# camera_url      =  0 # Connect to local webcam
```

--------------------------------

### Inspect Model Bundle CLI

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/edge-torch-models/README.md

Use this script to inspect published model bundles. It supports detailed analysis of models and state dictionaries, and works with both .pth and .pt files.

```bash
python inspect_model_bundle.py --bundle_path /path/to/published_model_bundle
```

```bash
python inspect_model_bundle.py --bundle_path /path/to/bundle --model_name qat_eval-saved_model
```

```bash
python inspect_model_bundle.py --bundle_path /path/to/bundle --detailed --show_weights
```

```bash
python inspect_model_bundle.py \
    --bundle_path /path/to/published_model_bundle \
    --model_name qat_eval-saved_model \
    --detailed
```

```bash
python inspect_model_bundle.py \
    --bundle_path /path/to/published_model_bundle \
    --detailed \
    --show_weights
```

--------------------------------

### Initialize Predictor Object

Source: https://github.com/landing-ai/landingai-python/blob/main/examples/apps/zoom-app/Zoom_Demo.ipynb

Create a Predictor instance using your endpoint ID and API key to interact with your deployed model.

```python
# Build the predictor object for your model
model = Predictor(endpoint_id, api_key=api_key)
```

--------------------------------

### Capture and Resize Webcam Frames to Video

Source: https://github.com/landing-ai/landingai-python/blob/main/docs/image-acquisition/webcam.md

Use this snippet to capture frames from your local webcam, resize them, and save them into a new video file. Ensure you have the necessary imports.

```python
from landingai.pipeline.image_source import Webcam
from landingai.pipeline.frameset import FrameSet

frameset = FrameSet() # (1)!
with Webcam(fps=1) as camera: # (2)!
    for i, frame in enumerate(camera): # (3)!
        if i >= 100:
            break
        frame.resize(width=256) # (4)!
        frameset.append(frame) # (5)!
frameset.save_video("/tmp/resized-video.mp4") # (6)!
```

--------------------------------

### Create LandingAI Frame from Image File

Source: https://context7.com/landing-ai/landingai-python/llms.txt

Create a Frame object from an image file path. Supports common image formats. Can also create from local or remote URIs.

```python
from landingai.pipeline.frameset import Frame
import numpy as np

# Create a Frame from an image file (supports jpg, png, bmp, etc.)
frame = Frame.from_image("/path/to/image.jpg")
print(f"Image size: {frame.image.size}")

# Create a Frame from a local or remote URI
frame = Frame.from_image("https://example.com/image.png")

# Create a Frame from a numpy array (e.g., from OpenCV)
# By default assumes BGR format (OpenCV standard)
img_array = np.random.randint(0, 255, (480, 640, 3), dtype=np.uint8)
frame = Frame.from_array(img_array, is_bgr=True)

# Create a Frame with custom metadata
frame = Frame.from_image(
    "/path/to/image.jpg",
    metadata={"camera_id": "cam_01", "timestamp": "2024-01-15T10:30:00"}
)
print(frame.metadata)  # {'camera_id': 'cam_01', 'timestamp': '2024-01-15T10:30:00'}
```

--------------------------------

### Initialize SnowflakeNativeAppPredictor with Password

Source: https://context7.com/landing-ai/landingai-python/llms.txt

Initialize the SnowflakeNativeAppPredictor using Snowflake account credentials and a password for authentication. Ensure you have the necessary Snowflake account details and endpoint ID.

```python
predictor = SnowflakeNativeAppPredictor(
    endpoint_id="your-endpoint-id",
    snowflake_account="your_account.region",
    snowflake_user="your_username",
    snowflake_password="your_password",  # Or use private key below
    native_app_url="https://your-native-app-url.snowflakecomputing.app"
)
```