Hi, We are using the OAK-D Pro Wide(with OV972). And we would like to check the RGB Depth alignment like below link. RGB Depth alignment — DepthAI documentation | Luxonis But based on our test it seems not available its RGB-Depth Alignment function anymore with the OAK-D Pro Wide(with OV9782). Is our test result is right? and are there any way to make this RGB-Depth Alighment with the OAK-D Pro Wide(with OV972)? **** In the picture below, you can see that the angle of view of the depth image is narrower than that of the RGB image. There is a person at the edge of the RGB image, but in the Depth image the person is out of the field of view and cannot be seen. [URL=https://i.imgur.com/xj2V8or.png][IMG]https://i.imgur.com/xj2V8or.png[/IMG][/URL] [URL=https://i.imgur.com/aowc1S3.jpg][IMG]https://i.imgur.com/aowc1S3.jpg[/IMG][/URL] Best regards, Ryan.

Hi @"RyanLee"#p4891 , Could you please [provide full MRE,](https://docs.luxonis.com/en/latest/pages/support/) together with minimal script you are using? Thanks ,Erik

Hi Based on our test you can check the RGB-Depth angle of view difference by just running the demo code provided by Luxonis like below. How do you think? [URL=https://i.imgur.com/c3TEnQE.png][IMG]https://i.imgur.com/c3TEnQEh.png[/IMG][/URL] [URL=https://i.imgur.com/E95z622.png][IMG]https://i.imgur.com/E95z622h.png[/IMG][/URL] Best regards, Ryan.

Hi @"RyanLee"#p4925 , The OV9782 color image has much wider FOV than the depth (as mono cameras need to be rectified for the depth) - so that's expected. Thanks, Erik

Hi @"erik"#p4931 That means that the RGB Depth alignment is not available for OAK-D Pro wide OV9782 anymore? **** https://docs.luxonis.com/projects/api/en/latest/samples/StereoDepth/rgb_depth_aligned/ Best regards, Ryan.

Hi @"RyanLee"#p4937 , It should be possible to rgb-detph align the model, but you would need [to undistort the rgb](https://github.com/luxonis/depthai-python/pull/661), as calibration uses undistorted rgb when calculating intrinsics. Thanks, Erik

HI @"erik"#p4945 We tried the code [undistort the rgb image](https://github.com/luxonis/depthai-python/blob/develop/examples/ColorCamera/rgb_undistort.py). But we got some error result like below.[URL=https://i.imgur.com/7u6DMs5.png][IMG]https://i.imgur.com/7u6DMs5.png[/IMG][/URL] Please check it out and let us know how to handle this issue. Best regards, Ryan.

Hi @"RyanLee"#p5141 , I have developed rgb-depth alignemnt demo that also does rgb undistortion; https://gist.github.com/Erol444/a9e50f5d8cef689dc86c544c15ca3175 I tried it with the OAK-D W and it works as expected. Thanks, Erik

HI @"erik"#p5143 I checked the test code you shared us above. It has been confirmed that the alignment between RGB/Depth has been significantly improved compared to before. However, the alignment between the two images still seems insufficient at the lens periphery for use in Spatial Detection. Is it possible to further improve the alignment between the two images more precisely? [URL=https://i.imgur.com/VPPudTd.jpg][IMG]https://i.imgur.com/VPPudTd.jpg[/IMG][/URL] [URL=https://i.imgur.com/a65HBtn.jpg][IMG]https://i.imgur.com/a65HBtn.jpg[/IMG][/URL] [URL=https://i.imgur.com/RmDbUYN.jpg][IMG]https://i.imgur.com/RmDbUYN.jpg[/IMG][/URL] Best regards, Ryan.

Hi @"RyanLee"#p5153 , Have you modified the script so it uses 800P color camera (which is max of OV9782), and potentially disabled isp scaling? So the first few lines of the script would be: ``` import cv2 import depthai as dai import numpy as np camRes = dai.ColorCameraProperties.SensorResolution.THE_800_P camSocket = dai.CameraBoardSocket.RGB ispScale = (1,1) ```

Hi @"erik"#p5162 This issue is cleared. So it is working now. And there was one of our mistake. https://gist.github.com/Erol444/a9e50f5d8cef689dc86c544c15ca3175#file-rgb-undistortion-depth-alignment-py Actually, this issue is cleared with above solution you shared. But we did some mistake (Bug) by ourselves. Thank you for your patience always. Best regards, Ryan.

OAK-D Pro wide OV9782 RGB-Depth Alignment

erik

Hi repena ,
Does this example now work as expected? https://docs.luxonis.com/projects/api/en/latest/samples/StereoDepth/rgb_depth_aligned/#rgb-depth-alignment

repena

Hello erik,

No, when I run this example it does not work as expected.

As you can see in this picture the whiteboard depth mask is unaligned with the rgb image, this can be clearly seen in the right hand-side of the picture.

Here is another example:

erik

Hi repena ,
Could you try this script?

repena

erik I tried this code but this is the output:

TestCA

erik

I have tried and added Image parameter like lumianace and few more computer vision algorthims for smoothing :

Btw how to reach to your team ?

erik

Hi repena , it seems like you don't have the newest calibration - are you using an older device? If that's the case, you can recalibrate the device.

repena

Hello Erik,

I am using an OAKD PRO Wide camera, as stated by the documentation there is no need to calibrate this camera since it is a new model.

Do you think I should still try to calibrate it?

jakaskerl

Hi repena
Yes. Shouldn't take to long and it could fix the issue you are having.

Thanks,
Jaka

LeonelSimo

Hi erik,

I am also using the wide camera (OV9782), and I have tried your script, but I have a delay of 3 seconds. How to fix it?

erik

Hi LeonelSimo , please see docs here: https://docs.luxonis.com/projects/api/en/latest/tutorials/low-latency/

jakaskerl

LeonelSimo
You can pass the --help to the calibrate script to see possible arguments. The capture delay can be set using: -cd <n seconds>.

Thanks,
Jaka

LeonelSimo

jakaskerl

It is possible to use the calibrate.py only for ser the capture delay? I can set for example 30ms or 60 ms, or what value do you suggest?

jakaskerl

Hi LeonelSimo
I'm not sure I understand the question. I was referring to the time delay it takes from pressing the capture button to when the calibration image is taken. There is no real reason but your own convenience. I'd set the delay to 0 so I don't have to wait.
Why those exact values (30ms, 60ms)?

Thanks,
Jaka

TestCA

Hi , I have tried with computer vision by python it can be enhanced by luminance and few more parameters :

jakaskerl

Hi @TestCA
Easiest way is to create a new thread here or write an email to support@luxonis.com if the matter is non-technical.

Thanks,
Jaka

« Previous Page