What is the optimal resizing images for process on limit resources?

I am testing WebODM to find about optimal resizing images that well-fit to processing resources.
The environment was set to run on local PC and cloud’s VM. Both use Docker’s container.
-PC specs: CPU: I7-4720HQ , 16 GB of memory and 100 GB storage
-VM on cloud : 4 cores CPU , 16 GB of memory and 100 GB storage
It is good if anyone please suggest optimal resolution of resizing parameter that will give a good image quality and processing time. thank you.

1 Like

What is your projected project size (image count and MP of images?)

1 Like

In my experiment. I have used 5472x3648 pixels from Phantom 4 RTK as input. There are 4 types of image set :fields, buildings, wasted land and mixed. Image set is up to 30-70 images per set . The drone fly at 60m(1.64cm/px), 90m(2.47cm/px) and 120m(3.29cm/px). Then I have tried resize them half by half (2736x1824 =50% and 1368x912 = r25% of raw 5472x3648) and monitor processing time, CPU, RAM, Storage being used and final orthomosaic quality. And later, compare each image size to evaluate which one is optimal size to fit provided resources.

2 Likes