Do I need Opencv for coding for realsense or is the sdk/c++ cover everything? I previously used c# with the kinect sdk and used the native kinect mapping functions(from rgb to depth and back) and the c# writable bitmap with no 3rd party product.

[ARoth4]

I ask because for some reason opencv is not working with my VS2017 setup(idk why) and I'd rather not even figure it out if I don't need to.

I ran the the sample code in the realsense sdk with VS and it runs smooth and the sdk functions are simple and seem to work.

When I google the d435 I get a lot of opencv though.

[MartyG]

OpenCV is fully supported by the RealSense SDK 2.0 software but is totally optional. You can write only in C# if that is the language that you are most comfortable with, though the number of example programs available to refer to when learning SDK programming is currently less than with C++.

[ARoth4]

Thanks. I'm fine with c++. It's faster anyway. Are there examples outside the sdk that are useful? All I've seen when i search are realsense api tutorial and then some opencv.

One more thing - the API has a "laser power" utility which I am unfamiliar with. Is that a niche utility or is it something I should know about?

Thanks.

[MartyG · Accepted solution]

The SDK has an index page of examples.

https://github.com/IntelRealSense/librealsense/tree/master/examples

The majority of scripting tends to be posted by RealSense users on the RealSense GitHub site. So a good web-search trick for googling for code examples is to include the word 'librealsense' as the first word in the search, as this is the word that search engines tend to index the GitHub's posts with. Example: 'librealsense align depth color c++'

There are also a lot of RealSense 400 Series related articles about products and tools outside of Intel on the link below.

https://forums.intel.com/s/question/0D50P0000490WyoSAE/realsense-stories-you-may-have-missed?language=en_US

The 400 Series cameras have an "IR Emitter" projector component that can project a pattern of semi-random dots onto the objects in a room in order to make low-detail textures such as walls and doors more easily analyzable for depth detail. The 'Laser Power' setting can decrease the visibility of this pattern by reducing the Laser Power value if you do not want the dots to be strongly visible on the camera image. Or you can turn the projection off completely.

[ARoth4]

Thanks. That's very useful information. I'm saved a copy your linked post.

I'll have to look into the laser. You wrote that it's for low detail textures. I suppose it returns some table data(per frame) that I can use in conjunction with the standard depth table? I'll look into it.

Thanks.