When I get an error when I use deepstream's NvDsDisplayMeta, I don't know where the problem is

1017948396 · April 3, 2023, 7:19am

error :corrupted size vs. prev_size while consolidating Aborted (core dumped)

mycode:

static GstPadProbeReturn
nvinfer_src_pad_buffer_probe1 (GstPad * pad, GstPadProbeInfo * info,
gpointer u_data)
{

GstBuffer *buf = (GstBuffer *) info->data;
NvDsBatchMeta *batch_meta = gst_buffer_get_nvds_batch_meta (buf);
NvDsMetaList * l_frame = NULL;

NvDsMetaList * l_obj = NULL;
NvDsObjectMeta *obj_meta = NULL;

// Get original raw data
GstMapInfo in_map_info;
if (!gst_buffer_map (buf, &in_map_info, GST_MAP_READ)) {
g_print (“Error: Failed to map gst buffer\n”);
gst_buffer_unmap (buf, &in_map_info);
return GST_PAD_PROBE_OK;
}

char* src_data = NULL;
NvBufSurface *surface = (NvBufSurface *)in_map_info.data;
int conunt=0;
for (l_frame = batch_meta->frame_meta_list; l_frame != NULL;
l_frame = l_frame->next)
{
//图像转换
NvDsFrameMeta *frame_meta = (NvDsFrameMeta *) (l_frame->data);
if (!frame_meta)
continue;

  guint height = surface->surfaceList[frame_meta->batch_id].height;
  guint width = surface->surfaceList[frame_meta->batch_id].width;
  cudaMalloc((void**)&src_data,surface->surfaceList[frame_meta->batch_id].dataSize);
  cudaMemcpy((void*)src_data,(void*)surface->surfaceList[frame_meta->batch_id].dataPtr,surface->surfaceList[frame_meta->batch_id].dataSize,cudaMemcpyDeviceToDevice);  
  cv::cuda::GpuMat nv12_mat=cv::cuda::GpuMat(height, width, CV_8UC4, src_data,surface->surfaceList[frame_meta->batch_id].pitch);
  cv::cuda::GpuMat nv12_mat_out=cv::cuda::GpuMat(height, width, CV_8UC4);
  cv::cuda::resize(nv12_mat, nv12_mat, cv::Size(512,512 ));
  cv::Rect area_new = cv::Rect(512, 0,512, 512);
  nv12_mat.copyTo(nv12_mat_out(area_new));
  cudaMemcpy((void*)surface->surfaceList[frame_meta->batch_id].dataPtr,(void*)nv12_mat_out.data,surface->surfaceList[frame_meta->batch_id].dataSize,cudaMemcpyDeviceToDevice);
  
  ///
  //cv::Mat cpu_Mat;
 // nv12_mat_out.download(cpu_Mat);
  //
  if(src_data != NULL)
  {
    cudaFree(src_data);
    src_data = NULL;
  }
  //图像转换 ---
  
  //渲染的设置  
  NvDsDisplayMeta *disp_meta = NULL;
  if (!disp_meta) {
     disp_meta = nvds_acquire_display_meta_from_pool(batch_meta);
     disp_meta->num_lines = 0;
  } //申请空间
    //首先画一个固定的框图  限定一个区域
  disp_meta->line_params[disp_meta->num_lines].x1 =0;
  disp_meta->line_params[disp_meta->num_lines].y1 =0;
  disp_meta->line_params[disp_meta->num_lines].x2 =512;
  disp_meta->line_params[disp_meta->num_lines].y2 =0;
  disp_meta->line_params[disp_meta->num_lines].line_width = 4;
  disp_meta->line_params[disp_meta->num_lines].line_color.red = 1.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.green = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.blue = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.alpha = 0.5;
  disp_meta->num_lines=disp_meta->num_lines+1;
  

  disp_meta->line_params[disp_meta->num_lines].x1 =0;
  disp_meta->line_params[disp_meta->num_lines].y1 =0;
  disp_meta->line_params[disp_meta->num_lines].x2 =0;
  disp_meta->line_params[disp_meta->num_lines].y2 =512;
  disp_meta->line_params[disp_meta->num_lines].line_width = 4;
  disp_meta->line_params[disp_meta->num_lines].line_color.red = 1.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.green = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.blue = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.alpha = 0.5;
  disp_meta->num_lines=disp_meta->num_lines+1;
  
  
  disp_meta->line_params[disp_meta->num_lines].x1 =512;
  disp_meta->line_params[disp_meta->num_lines].y1 =0;
  disp_meta->line_params[disp_meta->num_lines].x2 =512;
  disp_meta->line_params[disp_meta->num_lines].y2 =512;
  disp_meta->line_params[disp_meta->num_lines].line_width = 4;
  disp_meta->line_params[disp_meta->num_lines].line_color.red = 1.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.green = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.blue = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.alpha = 0.5;
  disp_meta->num_lines=disp_meta->num_lines+1;
  

  disp_meta->line_params[disp_meta->num_lines].x1 =0;
  disp_meta->line_params[disp_meta->num_lines].y1 =512;
  disp_meta->line_params[disp_meta->num_lines].x2 =512;
  disp_meta->line_params[disp_meta->num_lines].y2 =512;
  disp_meta->line_params[disp_meta->num_lines].line_width = 4;
  disp_meta->line_params[disp_meta->num_lines].line_color.red = 1.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.green = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.blue = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.alpha = 0.5;
  disp_meta->num_lines=disp_meta->num_lines+1;

  for (l_obj = frame_meta->obj_meta_list; l_obj != NULL;l_obj = l_obj->next)
  {
  obj_meta = (NvDsObjectMeta *) (l_obj->data);
  if (!obj_meta)
    continue;
    
  for (NvDsMetaList * l_user = obj_meta->obj_user_meta_list; l_user != NULL;
      l_user = l_user->next) 
   {
    NvDsUserMeta *user_meta = (NvDsUserMeta *)l_user->data;
    if(user_meta->base_meta.meta_type == (NvDsMetaType) NVDS_USER_JARVIS_META_PYVA) 
    {
     NvDsPyvaMetaData *PYVA_meta =
          (NvDsPyvaMetaData *) user_meta->user_meta_data; 
                   
    conunt=conunt+1;
    //大致的渲染结果      
    float x_center=PYVA_meta->x_center;
    float y_center=PYVA_meta->y_center;
    float width=PYVA_meta->width;  //0
    float height=PYVA_meta->height;//1
    
    
    float xtl=256-(y_center-height/2)/80*512;//x0  
    float ytl=341-(x_center-width/2)/120*512;//y0 
    
    
    
    float xbr=256-(y_center+ width/2)/80*512;//x3
    float ybr=341-(x_center+height/2)/120*512;//y3
    
    
    float rect_width=xbr-xtl;
    float rect_height=ybr-ytl;
    
     
    float x1=xtl+rect_width/2;
    float y1=ytl;
    
    
    float x2=xtl;
    float y2=ytl+rect_height/2;
      
  disp_meta->line_params[disp_meta->num_lines].x1 =0;
  disp_meta->line_params[disp_meta->num_lines].y1 =0;
  disp_meta->line_params[disp_meta->num_lines].x2 =100;
  disp_meta->line_params[disp_meta->num_lines].y2 =0;
  disp_meta->line_params[disp_meta->num_lines].line_width = 4;
  disp_meta->line_params[disp_meta->num_lines].line_color.red = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.green = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.blue = 1.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.alpha = 0.5;
  disp_meta->num_lines=disp_meta->num_lines+1;
  
  disp_meta->line_params[disp_meta->num_lines].x1 =100;
  disp_meta->line_params[disp_meta->num_lines].y1 =0;
  disp_meta->line_params[disp_meta->num_lines].x2 =100;
  disp_meta->line_params[disp_meta->num_lines].y2 =100;
  disp_meta->line_params[disp_meta->num_lines].line_width = 4;
  disp_meta->line_params[disp_meta->num_lines].line_color.red = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.green = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.blue = 1.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.alpha = 0.5;
  disp_meta->num_lines=disp_meta->num_lines+1;
  

  disp_meta->line_params[disp_meta->num_lines].x1 =0;
  disp_meta->line_params[disp_meta->num_lines].y1 =0;
  disp_meta->line_params[disp_meta->num_lines].x2 =0;
  disp_meta->line_params[disp_meta->num_lines].y2 =100;
  disp_meta->line_params[disp_meta->num_lines].line_width = 4;
  disp_meta->line_params[disp_meta->num_lines].line_color.red = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.green = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.blue = 1.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.alpha = 0.5;
  disp_meta->num_lines=disp_meta->num_lines+1;
  
  disp_meta->line_params[disp_meta->num_lines].x1 =0;
  disp_meta->line_params[disp_meta->num_lines].y1 =100;
  disp_meta->line_params[disp_meta->num_lines].x2 =100;
  disp_meta->line_params[disp_meta->num_lines].y2 =100;
  disp_meta->line_params[disp_meta->num_lines].line_width = 4;
  disp_meta->line_params[disp_meta->num_lines].line_color.red = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.green = 0.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.blue = 1.0;
  disp_meta->line_params[disp_meta->num_lines].line_color.alpha = 0.5;
  disp_meta->num_lines=disp_meta->num_lines+1;

     }
  }    
  
}

if(disp_meta)
nvds_add_display_meta_to_frame (frame_meta, disp_meta);
}
gst_buffer_unmap (buf, &in_map_info);
return GST_PAD_PROBE_OK;
}

yuweiw · April 3, 2023, 7:45am

Please provide complete information as applicable to your setup.

Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

1017948396 · April 3, 2023, 7:56am

Hardware Platform (Jetson / GPU)
Rtx3060
• DeepStream Version
Deepstream6.2
• TensorRT Version
Tensorrt 8.5.2
• NVIDIA GPU Driver Version (valid for GPU only)
525.85.12
• Issue Type( questions, new requirements, bugs)
Problems with using NvDsDisplayMeta.
**• How to reproduce the issue ?
It is rendered in a way similar to the drawing of lines for human key detection(deepstream_tao_apps/apps/tao_others/deepstream-bodypose2d-app at master · NVIDIA-AI-IOT/deepstream_tao_apps · GitHub)。
**• Requirement details。
The engineering that imitates the key points of the human body, the meta data of the rotating box is customized, the rendering part of the model inference, and the memory leakage problem occurs with the use of NvDsDisplayMeta.

yuweiw · April 3, 2023, 10:55am

Could you attach more log info with GST_DEBUG=3?

1017948396 · April 3, 2023, 11:01am

There are no mistakes when I draw 3 lines.

yuweiw · April 4, 2023, 1:47am

There is a maximum limit of num_lines，have you done any special process? You can refer to below:
https://github.com/NVIDIA-AI-IOT/deepstream_tao_apps/blob/master/apps/tao_others/deepstream-bodypose2d-app/deepstream_bodypose2d_app.cpp#L237

1017948396 · April 4, 2023, 1:52am

That’s it. Thank you. Problem solved。

system · April 18, 2023, 1:52am

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.

Topic		Replies	Views
How to NvBufSurface to cv::Mat? DeepStream SDK	12	460	July 2, 2024
[REOPEN] Nvstreamdemux does not copy obj_meta parent structure to src pad DeepStream SDK deepstream	39	1461	February 26, 2024
Issue with Cropping and Saving a Specific Frame Area in .jpg Format DeepStream SDK jetson , deepstream	18	73	March 10, 2025
Gst-dsexample error DeepStream SDK camera	6	742	November 2, 2022
Unable to add pgie_src_pad from deepstream-image-meta-test to deepstream-app DeepStream SDK	13	1189	December 29, 2021
Need some tips on debugging deepstream-app-based C++ application DeepStream SDK gstreamer	13	1929	June 1, 2022
Memory leak even after calling NvBufSurfaceUnMap DeepStream SDK deepstream	16	69	March 20, 2025
Memory Compatibility Error:Input surface gpu-id doesnt match with configured gpu-id for element DeepStream SDK	6	1319	November 29, 2021
Deepstream-test3 deepstream-5.0 DeepStream SDK ubuntu , gstreamer	12	1418	October 12, 2021
Unknown key 'blur-objects' for group [ds-example] DeepStream SDK	4	968	October 12, 2021

When I get an error when I use deepstream's NvDsDisplayMeta, I don't know where the problem is

Related topics