search

dme-compunet / YoloV8

Integrate YOLOv8 into your C# project for a variety of real-time tasks including object detection, instance segmentation, pose estimation and more, using ONNX Runtime.
https://www.nuget.org/packages/YoloV8
GNU Affero General Public License v3.0
248 stars 48 forks source link

Real-time Processing of MJPEG Video Stream using YOLOv8 #72

Open areytechai opened 3 days ago

areytechai commented 3 days ago

Hi,

I would like to know if it's possible to process the video stream from this MJPEG URL (video stream) in real-time using the YOLOv8 architecture.

Currently, what I'm doing is saving the video frames as JPEG images via VLC, then processing them with YOLOv8 and saving the results. However, this method isn't real-time.

Is there a way to process the video stream directly in real-time? If not, could you suggest some simple alternative methods or best practices to achieve this?

Thanks in advance for your help!

dme-compunet commented 3 days ago

I have no experience with this, but I guess the best way is to extract the frames with ffmpeg into the stdout and then process them with YOLOv8

areytechai commented 3 days ago

I have no experience with this, but I guess the best way is to extract the frames with ffmpeg into the stdout and then process them with YOLOv8

Thank you! Extracting frames with FFmpeg into stdout and then processing them with YOLOv8 is indeed an effective approach. By the way, when do you plan to incorporate YOLOv11? Additionally, we would like to count the bags being filled at the cat food factory using the camera. Is this possible?

dme-compunet commented 1 day ago

By the way, when do you plan to incorporate YOLOv11?

In general, once Ultralytics supports exporting it to onnx, then I can add support for it, But it's not guaranteed.

Additionally, we would like to count the bags being filled at the cat food factory using the camera. Is this possible?

Better to ask here: https://github.com/ultralytics/ultralytics