Repository navigation
Openvino sample - #5283
Openvino sample#5283
Conversation
* Simple face detection and depth reporting * Depends on BUILD_OPENVINO_EXAMPLES * Requires INTEL_OPENVINO_DIR and OpenCV_DIR be defined * Model files automatically downloaded in CMake
| STATUS status) | ||
| list(GET status 0 error_code) | ||
| if (NOT ${error_code} EQUAL 0) | ||
| message(FATAL_ERROR "Download TM2 firmwnare (${status}) - ${url}") |
There was a problem hiding this comment.
message(FATAL_ERROR "Error (${status}) downloading: ${url}")
| This example demonstrates OpenVINO™ toolkit integration with facial detection, using | ||
| basic depth information to approximate distance. | ||
|
|
||
| <p align="center"><img src="rs-face-vino.jpg" alt="screenshot"/></p> |
|
|
||
| ### Depth estimation | ||
|
|
||
| Depth is arrived at very simplistically: the center coordinates of each face on |
There was a problem hiding this comment.
is arrived the right word?
There was a problem hiding this comment.
Changed this section, since we now align
| // Wait for the next set of frames | ||
| auto data = pipe.wait_for_frames(); | ||
| // Make sure the frames are spatially aligned | ||
| //data = align_to.process( data ); |
| See the Windows-specific OpenVINO guide [here](https://docs.openvinotoolkit.org/latest/_docs_install_guides_installing_openvino_windows.html). | ||
|
|
||
| > NOTE: the code here will change for you and is given only as an example. | ||
| > Your experience may vary! |
There was a problem hiding this comment.
Removed that line... not sure what the intent was
| than specify a hard-coded path, they usually expect these to be available in the | ||
| same directory as the .exe. If the .exe is at `c:/work/lrs/build/Release` then | ||
| place any missing DLLs or models in the same directory, or change the code to | ||
| look in a different place. |
There was a problem hiding this comment.
Should this be automated somehow? Script?
There was a problem hiding this comment.
Added:
For example, the following OpenCV DLLs were needed for the non-Debug executable,
and were copied from ${OpenCV_DIR}/bin to the RelWithDebInfo directory where
the .exe is written:
opencv_core411.dll
opencv_highgui411.dll
opencv_imgcodecs411.dll
opencv_imgproc411.dll
opencv_videoio411.dll
The pre-trained model files used by the examples are provided automatically when
CMake is run, and are placed in each sample's build directory. For example, the
following three are placed in build/wrappers/openvino/face:
README.txt
face-detection-adas-0001.xml
face-detection-adas-0001.bin
Pre-trained models were taken from the OpenVINO model zoo,
and public models
were converted using the Model Converter.
* moved common CMake functionality (downloads, OPENVINO_FILES, ELPP_FILES) into openvino CMakeLists.txt * now aligns the depth and color frames * removed some not-useful debug messages from rs-vino/* * added labels to openvino_helpers::detected_face
rs-face-vino Sample
Overview
This example demonstrates OpenVINO™ toolkit integration with facial detection, using
basic depth information to approximate distance.