After acquiring display meta from the pool and filling it's text params, not able to see it on scree...

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Hello everyone,

I would like to ask that, I want to display the count text on each frame.

Therefore in my custom plugin, I have added the display meta to the frame meta after populating the display meta structure with the text i want. However, frame is rendered at the end without my added display meta for the text.

My current logic to achieve that is mentioned below:

In my gst-plugin which is of GSTBaseTransform type , I am reading each framemeta present in a batch which is being received in my plugin’s transform_ip () function.

My logic of adding the display meta to each frame is :

static void
attach_displaymetadata_full_frame (GstMyCount * mycount, NvDsFrameMeta * frame_meta, guint Source_id,
gdouble scale_ratio, guint batch_id)
{
NvDsBatchMeta *batch_meta = frame_meta->base_meta.batch_meta;
NvDsDisplayMeta *display_meta = NULL;
display_meta = nvds_acquire_display_meta_from_pool(batch_meta);

nvds_acquire_meta_lock (batch_meta);

  NvOSD_TextParams & text_params = display_meta->text_params[0];
  
  text_params.x_offset = 500;
  text_params.y_offset = 500;
   // Font face, size and color
  text_params.font_params.font_name = (char *)"Serif";
  text_params.font_params.font_size = 11;
  text_params.font_params.font_color = (NvOSD_ColorParams) {1, 1, 1, 1};
  // display_text required heap allocated memory
  text_params.display_text = g_strdup ("Count");

// Set black background for the text
text_params.set_bg_clr = 1;
text_params.text_bg_clr = (NvOSD_ColorParams) {0, 0, 0, 1};

nvds_add_display_meta_to_frame( frame_meta, display_meta);

nvds_release_meta_lock (batch_meta);

}

Notice: data is getting added to the framemeta(checked) but not being rendered on the display.

Looking forward to get some help from the nvidia’s deepstream library maintainers, as to what am i missing here.

2

Could you refer deepstream_test1_app.c->osd_sink_pad_buffer_probe?

3

One guess is maybe the coordinate out of the on screen window size, how about your output resolution?

4

Thanks bcao for pointing me towards the right direction, but the actual issue was : As after acquiring display_meta from the meta pool of the batch_meta, one needs to set display_meta->num_labels = 1; , as before filling in the display meta’s text_params for the entire frame, one should specify how many string labels will it be displaying beforehand.

Thanks

2 Likes