Align depth to color from saved images

[Todibo99]

---

Required Info
Camera Model	{ D415 / D435 }
Operating System & Version	{Ubuntu 18}
Kernel Version (Linux Only)	(e.g. 4.14.13)
Platform	PC
Language	{python }

Issue Description

Hi,
I have some saved color and depth images from D415 and D435. I would like to align depth to color. However, the pyrealsense2.align API requires a composite frame as input. I am wondering is there a way to restore frames from saved images or do you have any other suggestions?

Thanks!

[MartyG-RealSense]

Hi @Todibo99 When a depth frame is saved to PNG, most of the depth information is lost. It is therefore recommended to save depth to a format that preserves the depth information, such as a .raw image.

The link below features a Python script by a RealSense user that took the approach of aligning PNG color frames and an .npy file of the depth as an array of scaled matrices.

#4934 (comment)

An approach to the composite frame issue may be to manually create a custom frameset using software_device or a custom processing block.

#5847

Information about software_device can be found in the documentation of the SDK's C++ example program for it.

https://github.com/IntelRealSense/librealsense/tree/master/examples/software-device

The number of references available for using software_device with Python are more limited, though the link below may be helpful.

#4197

[Todibo99]

thx, i have sorted it out

[MartyG-RealSense]

Great to hear, @Todibo99 - thanks for the update!

[MartyG-RealSense]

Case closed due to solution achieved and no further comments received.