[Realsense SDK] Merging Scan & Align Object

[AHoan1]

Dear,

I'm working on Realsense SDK and SR300,

1 - How to merge two PXCM3DScan or create multi PXCM3DScan in a PXCMSenseManager;

2 - Having two parts of an object, how to auto align?

3 - In PXCM3DScan, can you show me how to Reconstruct 2 PXCM3DScan to a file?

4 - Having *rssdk files of a model, how to merge?

Thanks!

[AHoan1]

Dear,

Could you show me the conditions are the best that I scan a model 3D body full of R200 (range, light, configure in source code...)?

[MartyG]

The link below has a script for the old RSSDK RealSense SDK that combines RGB and depth streams into one image. The script was for the SR300's direct predecessor the F200.but should work with SR300.

https://mtaulty.com/2015/04/16/m_15794/ https://mtaulty.com/2015/04/16/m_15794/

Regarding the best conditions for scanning a 3D model with the R200: it is not something that can be described in words precisely, because each person's location is different. In general, you should scan in a location that has average lighting (not too dark, not too bright), avoid rooms with fluorescent lights like ceiling strip lights, and position the camera so that the object / person you are scanning fills as much of the camera's view as possible.

[AHoan1]

I want to merge 2 rssdk files not merge RBG and Depth.

[MartyG]

Sorry, when you said RSSDK, I thought you meant the old "RSSDK" SDKs. I do not have a clear answer for how you might merge two .RSSDK files. I can only offer this link to the last time that question was asked.

[AHoan1]

Dear,

Could you show me the conditions are the best that I scan a model 3D body full of SR300 (range, light, configure in source code...)?

Currently I set range is 60cm.

[MartyG]

The same rules for getting good scan results that I mentioned earlier for R200 apply to scanning with SR300 too.

The SetIVCAMFilterOption instruction can be used in the 2016 R2 and 2016 R3 SDKs to re-balance the output of F200 and SR300 cameras according to accuracy and distance.

https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/index.html?setivcamfilteroption_device_pxccapture.html Intel® RealSense™ SDK 2016 R2 Documentation

[AHoan1]

Dear,

How I can change configure for area of camera?

I use SDK R2 to scan model, but height is limited.

[MartyG]

If you are scanning a human body, try clicking on the arrow at the end of the Full option to drop the menu down, and select the 'Body' option.

[AHoan1]

I know, but the option 'BODY' can not scanning.

Could you please try scan with option 'BODY' and take a capture for me.

I have tried so many times before and failed.

Please help me change configure of are.

[MartyG]

Reading the 3DScan instructions, it looks like the Full option was the correct one for a full human body, as it provides the largest scanning area.

If the camera cannot see the whole body then you should move the camera further away from the body so the camera can see more of the body.

[AHoan1]

Code sample of intel is a deception.

[AHoan1]

I use SR300, if depth over 50cm, it is out of range?

[MartyG]

Sorry, I forgot you were using an SR300 and not a R200 with much longer range. The SR390 is not suited to scanning the full human body because of its limited range. The R200 is the best camera for full body scanning. Because of the SR300's range, it is better at just scanning the head.

The only way I can think of to get depth range on the SR300 like that of the R200 is to use IVCamFilterOption to give the SR300 longer range at the expense of scanning accuracy.

https://software.intel.com/sites/landingpage/realsense/camera-sdk/v2016r3/documentation/html/index.html?ivcamfilteroption_device_pxccapture.html Intel® RealSense™ SDK 2016 R3 Documentation

[AHoan1]

Dear MartyG,

Have a nice day,

When scanning a model, how to catch "lost tracking"?

I tried:

// Subscribe to receive range and tracking alerts

scan.Subscribe(OnAlert);

It does not seem to work.

[AHoan1]

How to check the camera does not work?

[MartyG]

The 3D scanning function of the 2016 R2 SDK's AlertEvent system has a type of alert called ALERT_LOST_TRACKING

https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/index.html?alertevent_pxc3dscan.html Intel® RealSense™ SDK 2016 R2 Documentation

The script below shows an example of use of AlertEvent.

https://github.com/getnamo/realsense-ue4/blob/master/Plugins/RealSensePlugin/ThirdParty/RSSDK/Include/pxc3dscan.h realsense-ue4/pxc3dscan.h at master · getnamo/realsense-ue4 · GitHub

[AHoan1]

Thank you,

I did it. I have a question:

How to check the camera does not work?

Could you show me sample source code?

[MartyG]

If you type 'camera explorer' into the text box at the bottom of your Windows screen (next to the button where you shut down the PC) then you can find the Camera Explorer program. This checks which RealSense cameras you have plugged in, if they can be detected, and lets you do tests with them.

[AHoan1]

No! That's not what I meant.

When scanning a model, camera SR300 is got stuck.

It does not work.

How to catch that case in the source code?

[MartyG]

In the 2016 R2 SDK, AcquireFrame can be used to check if a stream has 'timed out' because it has got stuck.

https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/index.html?acquireframe_pxcsensemanager.html Intel® RealSense™ SDK 2016 R2 Documentation

[AHoan1]

Dear,

Why is The Camera SR300 scanning then get stuck?

After I press start camera, about one-minute the camera does not work.