Storing the current frame in the bbox_generated_post_analytics_cb function in deepstream-app

[cy960518]

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) Jetson Xavier NX
• DeepStream Version DeepStream 6.0
• JetPack Version (valid for Jetson only) 4.6
• TensorRT Version 8.0.1-1
• Issue Type( questions, new requirements, bugs) questions

Is there any way that I could store the current frame as an image file in the “bbox_generated_post_analytics_cb” function in the deepstream-app?

What I want to do is whenever the YOLOv3 model detects new object, it runs the smart video record to record a short video, and store the current frame as a image file. I actually made my program to activate the smart video record whenever it detects a new object, however, have no idea how to make the program to get the current frame, and store it as a image file.

Thanks

[kesong]

Can you use nvds_obj_enc_process () in /opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/deepstream-image-meta-test?

[cy960518]

Well, seems like it is a little bit different.
What I understand by reading the codes in nvds_obj_enc_process() code in the /opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/deepstream-image-meta-test is that, this method simply encodes the enqueued object crops as JPEG.

What I actually want to do is encoding the whole frame.
Is it possible to do this with nvds_obj_enc_process() ? or am I missing something?

Thanks

[kesong]

Yes. make the whole image as one object.