Thanks fr your valuable support.
We have ordered ZR300 and we may pre-order D400 series as well.
1. We can't run sample applications without camera module.
Is it possible to pass image as an input to any sample application and detect the object/person ?
Is there any sample available to use an existing image as an input ?
If sample is not available, is it possible to code using "Intel realsense SDK for linux" ?
I want to use pre-made files instead of live capture.
1. what is API used to load pre-made files and catch the frame by frame?
2. Is it possible to use to detect/track person any general video/image files which captured using any other camera's ?
3. Is it possible to track Leg using Body Tracking library (Skeleton tracking)
4. In Body Tracking, it it possible to calculate angles, Length of hands / Legs ?
1. If you are using a ZR300, you should use the playback and record functions in the RealSense SDK For Linux.
For the 400 Series cameras, RealSense SDK 2.0 comes with a sample program called 'Save to Disk'. I've attached the text readme documentation file for this sample to this message.
2. Regarding reading person data from a non-RealSense recorded video, this discussion may be of use to you:
3. For leg tracking, I would recommend using OpenCV in combination with a 400 Series Camera and RealSense SDK 2.0.
4. In my own full-body project, I use the Unity game creation engine to move virtual objects representing RL leg joints using camera inputs and calculate the angles from the angles of the virtual objects instead of the angles of the RL leg joints.
I discuss the principles of my custom-built 'CamAnims' system in Unity in this article: