Table of Contents
As Artefact, we care about positively impacting people, the environment, and the community. That’s why we are committed to partnering with nonprofit organizations that make these values the basic building blocks of their vision.
Therefore, we collaborated with Smart Parks, a Dutch company that provides advanced sensor solutions to conserve endangered wildlife and efficiently manage park areas by providing cutting-edge technology.
The goal of the project was to use an ML approach to ingest the data coming from the camera traps and then provide insights, such as the presence of people or specific kinds of animals in the images or videos captured by the cameras. This information then is used by the park rangers to better protect the wildlife and sooner detect possible dangers like poachers.
- Cloud Storage - Used to store the camera trap media
- BigQuery - Used to store the API responses
- Cloud Vision API - Used for image analysis
- Cloud Video Intelligence API - Used for video analysis
- Cloud Run - Used to deploy the Web App
- Cloud Functions - Used as core component of the workflow
- Node-RED - Used to create events in Earth Ranger
- Streamlit - Used to develop the Web App
As you can see everything starts when an image or a video is taken by a camera trap. The device sends an email to a dedicated mailbox. A Node-RED flow is then used to check if there are any new emails if so, it uploads their attachments to a Google Cloud Bucket. The upload triggers the rest of the pipeline. The Cloud Function is executed, it calls Google Cloud Vision APIs and:
-
Saves the responses in BigQuery
-
Saves the annotated images/videos in another bucket in Cloud Storage
-
Sends the responses to a Node-RED flow
Finally, from Node-RED the responses are sent to Earth Ranger, a real-time software solution that aids protected area managers, ecologists, and wildlife biologists in making more informed operational decisions for wildlife conservation.
In addition, we developed a simple web app that reads from Cloud Storage and BigQuery and allows the users to manage and monitor the camera traps.
-
cloud function: This folder contains the complete cloud function code and all the instructions needed to recreate it in GCP
-
web app: This folder contains the complete streamlit app code and all the instructions needed to deploy it in Cloud Run
-
node red flows: This folder contains the JSON files of used the Node-RED flows and all the instructions needed to import them in your node-RED environment
Simone Gayed Said - simone.gayed-said@artefact.com
Project Link: https://github.com/artefactory/smartparks