Preprocess ROI information in object meta?

[stct]

• Hardware Platform: Lenovo Legion 7 16IAX7 with a NVIDIA GeForce RTX 3080
• DeepStream Version: 7.07.0.0
• TensorRT Version: 8.6.1.6
• NVIDIA GPU Driver Version (valid for GPU only): 12.2

Description

I use a NvPreprocess element to perform inference on multiple ROI per frame. In the same way as detections of objects at the frame border are clipped to the image border, detections at the ROI border are clipped to the ROI border as well. I would like to keep this for the moment. However, I would like to know which detection is related to which ROI to filter or merge those detections at the frame border if necessary. Is there any information attached to the object meta which I can access using the Python API?

[junshengy]

Add the below code in your python application, you can get more information through gst-inspect-1.0 nvinfer

nvinfer.set_property("clip-object-outside-roi", false)

@stct

Is there any information attached to the object meta which I can access using the Python API?

Generally speaking, yes, object meta includes bbox, label, etc.

https://docs.nvidia.com/metropolis/deepstream/dev-guide/python-api/PYTHON_API/NvDsMeta/NvDsObjectMeta.html

[yingliu]

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks