How to build AI image recognition and matching platform?

[roland423]

Select Topic Area

Question

Body

I'm building platform that managing image recognition and video analytics on GCP.
The data is captured from drones and want to implement this function in real-tiem.
I need help and I want to know if anyone has open-source.

[0x4k-helios]

That sounds like a great project! For real-time image recognition and video analytics, Google Cloud has tools that can help. You can use its services to stream data from your drones, process it in real-time, and apply machine learning models to analyze the images and videos. Tools like Pub/Sub and Dataflow are great for handling live data streams, and Vertex AI can help with building and deploying machine learning models.

For open-source solutions, you might want to check GitHub for pre-trained models or libraries for image recognition and video processing, like TensorFlow or OpenCV. These can save you a lot of time in development.

If you have specific challenges, feel free to share them, and I can help with more targeted advice!

[roland423]

Hi, @0x4k-helios
Thanks you for your help.
I also want to know if Google Cloud Platform Vertex AI supports real-time video analytic, according my research, the Vertex Ai does not support.

[0x4k-helios]

Well, actually Google Cloud Platform Vertex Ai does not support real-time video directly, so you have to use Google Cloud Pub/Sub or Cloud Storage, Dataflow for real-time analytic.