No Edge Data in Depth Image

[秦文翔]

Hello

I got another problem. Here are my depth images:

from the depth images above, you can see obvious black edge around the object. That means I can't get the edge depth data of an object.

Is it the shortage of intel realsense camera? Or is there some mistakes in my operation?

I want to measure a box. So I need the depth data of corners to get world coordinates of the corners.

I would be honored if you give me some suggestions about the problem or method of measurement.

[MartyG]

RealSense is capable of measuring boxes, as the Chinese retail giant JD.com used RealSense in their warehouse a few years ago to measure cardboard boxes. you can find more information on box measurement with RealSense, including YouTube videos, by googling for 'realsense measure box'.

The amount of black space around your body suggests that there is a strong light source where you are sitting that makes it hard for the camera to see that area. I saw this kind of black area recently when a camera was mounted on a ceiling near a ceiling light, and in the sky in an outdoor scene where the sun was shining.

Regarding the black outline around the box, I do not think that is related to the light source, given that the area around the box is okay. It is more likely that the camera is having trouble seeing the edges of the box, as flat or simplistic object edges are harder to see than complex edges that change direction a lot.

It is also possible that the room's light source may be reflecting off the edges of the box and making them more difficult to see.

[秦文翔]

Thank you very much!

I will search for other methods.

If you find some useful papers, please give me a link.

[MartyG]

In cases where you are having trouble generating a good quality depth image, another method other than adjusting the light source is to use a Point Cloud.

https://software.intel.com/en-us/articles/intel-realsense-technology-and-the-point-cloud Intel® RealSense™ Technology and the Point Cloud | Intel® Software

I see also that the image of the box is a bit diagonal, as though the camera is tilted a little. On one side of the box the black lines are thicker than on the other side. This could lend support to the theory of light reflecting off the edges, as the light reflections off the edges may behave differently on each side if the camera is not fully flat horizontal.