For a system that uses data from multiple cameras to create a single image, such as on some Huawei phones with multiple cameras, how does it deal with close subjects?
Cameras are only few tens of mm apart, but for close objects, the view from two cameras can be vastly different. How do you combine these two images for better noise level, etc?
I don't own such a phone. However, based on sample images available online, some phones (such as the Samsung S9+) store images from both cameras in the same JPEG file. See How to extract depth information from JPEG files?
With the Google Camera app, when taking pictures that contain depth information, there is a tool to select a different focus point. There is likely something similar to select which image is the primary image. The specific interface is likely to vary by phone.
The secondary camera of some phones captures only grayscale images. At the minimum focusing distance of the cameras, the main subject should be visible in both cameras. Parallax may be used to calculate a depth map. If the images cannot be correlated with each other, the secondary image is likely simply ignored.
