Why Do Pix4Dcloud Projects Fail During Point Cloud Generation on Large Image Sets?

PIX4Dcloud PIX4Dcloud Questions/Troubleshooting
1

Hello

I am facing repeated failures when trying to generate dense point clouds on Pix4Dcloud using large image sets (400+ images, high resolution). :upside_down_face:The initial steps (upload, calibration) complete successfully, but the processing stops during the point cloud generation phase with a vague error message. :slightly_smiling_face:

This doesn’t happen when running the same project locally using Pix4Dmapper. I have checked for image overlap and quality issues; but everything looks fine. :innocent:

Is there a known image count or resolution threshold for Pix4Dcloud that triggers processing failures? Or perhaps limits on RAM or processing time per project? :thinking: I would also like to know if anyone has successfully split large datasets for cloud processing / if there’s a smarter way to batch upload tiles without manual intervention. Cloud processing should ideally scale but right now; I’m hitting a wall.

I have checked What are the processing options - PIX4Dcloud related to this and found it very helpful.

While debugging this; someone on our team asked what is cloud architect and it made me realize that cloud infrastructure and workload orchestration are crucial even in photogrammetry platforms like Pix4Dcloud. :innocent:

Any tips or workflow adjustments from others who deal with large-volume aerial imagery would be really helpful.

Thank you !! :slightly_smiling_face:

2

Hello @benof

Welcome to Pix4D Community!

Is there a known image count or resolution threshold for Pix4Dcloud that triggers processing failures?

There is a limit to the maximum number of images that PIX4Dcloud can process: 4,000 images. However, depending on other factors such as image resolution, image content, etc., the actual number of images that PIX4Dcloud can process is lower.

The failure of datasets to process relative to the point cloud is often due to noise in the point cloud that exponentially increases the output size on the server, leaving it without space or RAM to complete this task. This happens especially when the processing area is not used, hence our insistence on its use.

In your particular case, we would need to review exactly what happened. Would you like to post a project URL here for us to review? Note that you can also open a support ticket and send us all the necessary information.

3

Pix4Dcloud projects can fail generating point clouds on large image sets due to high processing demands, limited server resources, or image quality issues. Splitting the project or reducing images can help.