RealSense SDK 2.0 documentation and papers can be found here:
The EnableObjectRecognition instruction, which is in the '2016 R2' RealSense SDK, is not in RealSense SDK 2.0. This is because the design of these two SDKs is completely different. In RealSense SDK 2.0, you may be able to achieve object detection / recognition by using SDK 2.0 in combination with the OpenCV software (which SDK 2.0 is fully compatible with) and a sample program such as the one below.
可以在此处找到RealSense SDK 2.0文档和论文：
EnableObjectRecognition指令位于“2016 R2”RealSense SDK中，不在RealSense SDK 2.0中。 这是因为这两个SDK的设计完全不同。 在RealSense SDK 2.0中，您可以通过将SDK 2.0与OpenCV软件（SDK 2.0完全兼容）和示例程序（如下面的示例程序）结合使用来实现对象检测/识别。
Thanks very much
I need sdk2.0's API document. Do you have a download address? Http://intelrealsense.github.io/librealsense/doxygen/md__c_1_jenkins_sys_rsbuild_workspace__l_r_s_doxygen_source_include_readme.html
I can't open the class pages in this document, probably because of the Chinese network.
What I need now is to look for something similar to the cylinder and calculate the volume of the cylinder. Can opencv find 3D objects? For example, can opencv be detected when the camera is overlooking the cylinder?
Here are some API documentation links for you.
Please try this link to the class list:
You could maybe get the volume of a cylinder without needing OpenCV if you adapt the 'Measure' sample program that measures objects to add a further calculation for the volume..First you would find the radius of the object and then multiply the square of its radius by pi, or 3.14 to get the cylinder volume
A YouTube video explains about calculating volume.
You would probably still need OpenCV though if you wanted to detect the presence only of cylinders and ignore other objects.
Those colored stripes are correct for the Measure sample. If you look at the image on the Measure sample's documentation page, it has those stripes across the image.
A member of the RealSense support team said today on the RealSense GitHub help forum about scanning at close distance: "Since our camera use pixel disparity to calculate the depth, if you want to reach to around 10cm using D435, you need to lower down the resolution to 424*240. With higher resolution, the minimum distance will be longer".