"PXC_" is undefined

[TCwh]

I am new to Intel RealSense products and the SDK. With my new camera D415, I am trying to create a new visual studio project on 3d scanning. First I tried to build and compile the samples given when I installed the SDK on my Windows 10, for example the pointcloud project with rs-pointcloud.cpp. The samples are running well with no errors. Then, to take a step further, I tried to build my first project by referencing the documentation: https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/index.html?doc_scan_coding_3d_scanning.html Intel® RealSense™ SDK 2016 R2 Documentation.

I was experimenting on Example 74 shows a typical 3D scanning program using SenseManager (Algorithm Modules> 3D Scan> Coding 3D Scanning). However, it was not working as PXCSenseManager was not found. In the doubt of not linking nor including the required dependecies correctly in my VS project, I copied the lines in Example 74 to the samples provided (in the installation directory, which is working). The same problem arises:

Meanwhile, most of the other examples given in the documentation seems to be using a lot of identifies with a "PXC-" prefix, but none of these can be used in my project. Upon a quick search on this community and google, I have found the reason is the "pxc_.h" headers are missing in my path. However, I could nowhere find such header files in my installation directory, nor the source codes I downloaded from your github repo https://github.com/IntelRealSense/librealsense GitHub - IntelRealSense/librealsense: Intel® RealSense™ SDK .

Is there anyone who knows how to solve this problem or where to find those headers, or who had also faced similar problems like me?

Thank you

[MartyG · Accepted solution]

Unfortunately, the '2016 R2' examples are not compatible with RealSense SDK 2.0. You will need to use examples that are based on SDK 2.0, which also has totally different instructions to the ones that 2016 R2 uses.

You can also find the example programs here:

https://github.com/IntelRealSense/librealsense/tree/master/examples librealsense/examples at master · IntelRealSense/librealsense · GitHub

Some of them are also in the SDK's folder, at Intel RealSense SDK 2.0 > Samples

For the documentation for SDK 2.0, I recommend using this link:

In regard to 3D scanning: at the moment it is not as easy with the 400 Series cameras as it was in 2016 R2. You need to create a point cloud depth scan and then convert it into a solid mesh. The link below may be helpful in explaining this.

[TCwh]

Thank you MartyG!

You have mentioned that the '2016 R2' examples are not compatible with RealSense SDK 2.0 anymore, does it mean that there is no official examples on, for example, face/hand recognition anymore, which is available in the '2016 R2' documentation? I have found many of the functions provided in that documentation quite useful and I would like to adopt those functions in my project. I tried to find their counterparts in the documentation SDK 2.0 documentation you provided, but it seems that is not.

Also, for the 3D scanning example you provided, I have to scan and output the point cloud with RealSense Viewer manually. Is there any way (maybe using the SDK) to automate the procedure?

One more thing, I am also consider coding my project in python. Is there any (official documentation) for the python version?

Thank you for your reply again.

[MartyG]

1. Here is a guide to creating point clouds with scripting in SDK 2.0.

https://github.com/IntelRealSense/librealsense/tree/master/examples/pointcloud librealsense/examples/pointcloud at master · IntelRealSense/librealsense · GitHub

2. For camera functions that go beyond basic depth analysis, Intel are encouraging companies to partner with them to provide those tools, and also encourage developers to submit their own code to SDK 2.0, which is open source. You can also integrate face and hand modules in the OpenCV software into your SDK 2.0 application, as SDK 2.0 is fully compatible with OpenCV.

An example of third-party face software that has been used with the 400 Series cameras is CloudCompare.

Skeletal tracking is also available for the 400 Series through the Nuitrack software.

https://www.youtube.com/watch?v=gMPtV4NXtUo Intel RealSense D415/D435 and Nuitrack skeletal tracking SDK replace Kinect SDK - YouTube

3. The documentation for the Python wrapper was just one page until recently.

https://github.com/IntelRealSense/librealsense/tree/master/wrappers/python librealsense/wrappers/python at master · IntelRealSense/librealsense · GitHub

However, In the last couple of weeks Intel have geared up their efforts to create new tutorial material, as they see doing so as the way forward in regard to educating about RealSense. The first of these new guides is for Python.

[TCwh]

Thank you MartyG

I think your links are helpful and please let me take time and study them first.

[MartyG]

You're very welcome. Please feel free to return to the forum any time. Good luck!