Segmentation fault on plotting rectangle on image

edit_or · January 30, 2024, 9:37am

Please provide complete information as applicable to your setup.

**• Hardware Platform (Jetson / GPU)**Orin Nano
• DeepStream Version 6.3
**• JetPack Version (valid for Jetson only)**5.1.3
I’m trying to plot rectangle on gst-dsexample plugin.
I have segmentation fault when cv::rectangle api is applied.
When segmentation fault happened for that API?

The whole code is as follows.

static GstFlowReturn
get_converted_mat (GstDsExample * dsexample, NvBufSurface *input_buf, NvOSD_RectParams * crop_rect_params, gint idx, vector<NvOSD_RectParams> &objects, gdouble & ratio, gdouble & ratio_w, gdouble & ratio_h)
{
  NvBufSurfTransform_Error err;
  NvBufSurfTransformConfigParams transform_config_params;
  NvBufSurfTransformParams transform_params;
  NvBufSurfTransformRect src_rect;
  NvBufSurfTransformRect dst_rect;
  NvBufSurface ip_surf;
#ifdef WITH_OPENCV
  cv::Mat in_mat;
#endif
  ip_surf = *input_buf;

  ip_surf.numFilled = ip_surf.batchSize = 1;
  ip_surf.surfaceList = &(input_buf->surfaceList[idx]);
  //file name
  std::time_t rawtime;
  std::tm* timeinfo;
  char buffer [80];
  std::time(&rawtime);
  timeinfo = std::localtime(&rawtime);
  std::strftime(buffer, 80, "%d%m%Y_%H%M_bishan_", timeinfo); 
  cnt++;
  std::stringstream stream;
  stream << cnt;
  std::string str;
  stream >> str;
  str = str+".jpg";  
  strcat(buffer, str.c_str()); 
                  
  gint src_left = GST_ROUND_UP_2((unsigned int)crop_rect_params->left);
  gint src_top = GST_ROUND_UP_2((unsigned int)crop_rect_params->top);
  gint src_width = GST_ROUND_DOWN_2((unsigned int)crop_rect_params->width);
  gint src_height = GST_ROUND_DOWN_2((unsigned int)crop_rect_params->height);
  
  double hdest = dsexample->processing_width * src_height / (double) src_width;
  double wdest = dsexample->processing_height * src_width / (double) src_height;
  guint dest_width, dest_height;

  if (hdest <= dsexample->processing_height) {
    dest_width = dsexample->processing_width;
    dest_height = hdest;
  } else {
    dest_width = wdest;
    dest_height = dsexample->processing_height;
  }
  
  transform_config_params.compute_mode = NvBufSurfTransformCompute_Default;
  transform_config_params.gpu_id = dsexample->gpu_id;
  transform_config_params.cuda_stream = dsexample->cuda_stream;

  err = NvBufSurfTransformSetSessionParams (&transform_config_params);
  if (err != NvBufSurfTransformError_Success) {
    GST_ELEMENT_ERROR (dsexample, STREAM, FAILED, ("NvBufSurfTransformSetSessionParams failed with error %d", err), (NULL));
    goto error;
  }

  ratio = MIN (1.0 * dest_width/ src_width, 1.0 * dest_height / src_height);

  if ((crop_rect_params->width == 0) || (crop_rect_params->height == 0)) {
    GST_ELEMENT_ERROR (dsexample, STREAM, FAILED, ("%s:crop_rect_params dimensions are zero",__func__), (NULL));
    goto error;
  }

#ifdef __aarch64__
  if (ratio <= 1.0 / 16 || ratio >= 16.0) {
    goto error;
  }
#endif
  src_rect = {(guint)src_top, (guint)src_left, (guint)src_width, (guint)src_height};
  dst_rect = {0, 0, (guint)dest_width, (guint)dest_height};
  /* Set the transform parameters */
  transform_params.src_rect = &src_rect;
  transform_params.dst_rect = &dst_rect;
  transform_params.transform_flag = NVBUFSURF_TRANSFORM_FILTER | NVBUFSURF_TRANSFORM_CROP_SRC | NVBUFSURF_TRANSFORM_CROP_DST;
  transform_params.transform_filter = NvBufSurfTransformInter_Default;

  /* Memset the memory */
  NvBufSurfaceMemSet (dsexample->inter_buf, 0, 0, 0);

  GST_DEBUG_OBJECT (dsexample, "Scaling and converting input buffer\n");

  /* Transformation scaling+format conversion if any. */
  err = NvBufSurfTransform (&ip_surf, dsexample->inter_buf, &transform_params);
  if (err != NvBufSurfTransformError_Success) {
    GST_ELEMENT_ERROR (dsexample, STREAM, FAILED, ("NvBufSurfTransform failed with error %d while converting buffer", err), (NULL));
    goto error;
  }
  /* Map the buffer so that it can be accessed by CPU */
  if (NvBufSurfaceMap (dsexample->inter_buf, 0, 0, NVBUF_MAP_READ) != 0){
    goto error;
  }
  if(dsexample->inter_buf->memType == NVBUF_MEM_SURFACE_ARRAY) {
    /* Cache the mapped data for CPU access */
    NvBufSurfaceSyncForCpu (dsexample->inter_buf, 0, 0);
  }

#ifdef WITH_OPENCV
  /* Use openCV to remove padding and convert RGBA to BGR. Can be skipped if
   * algorithm can handle padded RGBA data. */
  in_mat =
      cv::Mat (dsexample->processing_height, dsexample->processing_width,
      CV_8UC4, dsexample->inter_buf->surfaceList[0].mappedAddr.addr[0],
      dsexample->inter_buf->surfaceList[0].pitch);

#if (CV_MAJOR_VERSION >= 4)
  cv::cvtColor (in_mat, *dsexample->cvmat, cv::COLOR_RGBA2BGR);
#else
  cv::cvtColor (in_mat, *dsexample->cvmat, CV_RGBA2BGR);
#endif
#endif
   
   for(int obj_id=0; obj_id < objects.size(); obj_id++){
      //gint obj_left = GST_ROUND_UP_2((unsigned int)objects[obj_id].left);
      //gint obj_top = GST_ROUND_UP_2((unsigned int)objects[obj_id].top);
      //gint obj_right = obj_left + GST_ROUND_DOWN_2((unsigned int)objects[obj_id].width);
      //gint obj_bottom = obj_top + GST_ROUND_DOWN_2((unsigned int)objects[obj_id].height);      
      cv::Point p1(10, 10); 
      cv::Point p2(100, 100); 
      cv::rectangle(in_mat, p1, p2, cv::Scalar(0, 0, 255), 3, cv::LINE_8);
   }
   cv::imwrite(buffer, in_mat); 
   if (NvBufSurfaceUnMap (dsexample->inter_buf, 0, 0)){
      goto error;
   }

/*  if(dsexample->is_integrated) {
#ifdef __aarch64__
    
    if (USE_EGLIMAGE) {
      if (NvBufSurfaceMapEglImage (dsexample->inter_buf, 0) !=0 ) {
        goto error;
      }
      NvBufSurfaceUnMapEglImage (dsexample->inter_buf, 0);
    }
#endif
  }
*/ 
  return GST_FLOW_OK;

error:
  return GST_FLOW_ERROR;
}

yuweiw · January 31, 2024, 6:53am

Could you refer to our source code in gstdsexample.cpp?

/* rectangle for cropped objects */
  crop_rect = cv::Rect (crop_rect_params->left, crop_rect_params->top,
  crop_rect_params->width, crop_rect_params->height);

edit_or · January 31, 2024, 7:47am

That is cropping rectangle. I am plotting a rectangle on image using opencv’s api cv::rectangle

yuweiw · January 31, 2024, 10:59am

If you just want to draw the rectangle on the image, you don’t have to use the opencv or gstexample plugin. You can just use a probe function and set some parameters to our NvDsDisplayMeta like deepstream-test1.

edit_or · January 31, 2024, 11:02am

Ok let me check. Thanks

edit_or · February 1, 2024, 11:57am

My main intention is I like to plot rectangle on the saved image.
I have earlier posts like plotting bounding box on saved image, in which I tried to shift the position of gst-dsexample plugin to after osd plugin. But doesn’t succeed.
@ rajupadhyay59 suggested me to plot using opencv.
So I tried to plot rectangle using opencv and have faced another issue like Segmentation fault on plotting rectangle on image. Then you suggested me to use NvDsDisplayMeta.
My dificulty is to plot rectangle on the saved image.

This is my code in gst-dsexample to save image in gst_dsexample_transform_ip function.

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->class_id == PGIE_CLASS_ID_NOHELMET) {
                  found_nohelmet = true;
                  //objects.push_back(obj_meta->rect_params); 
                   NvDsObjEncUsrArgs frameData = { 0 };
		   frameData.isFrame = 1;
		   frameData.saveImg = true;
		   frameData.attachUsrMeta = false;
		   frameData.scaleImg = FALSE;
		   
		   std::time_t rawtime;
                  std::tm* timeinfo;
                  char buffer [80];
                  std::time(&rawtime);
                  timeinfo = std::localtime(&rawtime);
                  std::strftime(buffer, 80, "%d%m%Y_%H%M_bishan_", timeinfo); 
                  cnt++;
                  std::stringstream stream;
                  stream << cnt;
                  std::string str;
                  stream >> str;
                  str = str+".jpg";                 
                  strcpy(frameData.fileNameImg, buffer ); 
                  strcat(frameData.fileNameImg, str.c_str());
		   frameData.scaledWidth = 0;
		   frameData.scaledHeight = 0;
		   frameData.quality = 80;
	           nvds_obj_enc_process (dsexample->obj_ctx_handle, &frameData, surface, NULL, frame_meta);
	           queue.push(data_(frameData.fileNameImg, rawtime, obj_meta->object_id),  frame_meta->source_id);
              }
          }

yuweiw · February 2, 2024, 5:59am

The nvdsosd plugin draws the rectangle to the image. Is your dsexample plugin after the nvdsosd plugin in your pipeline?
There are many ways to draw the rectangle on the image. But if you don’t use the opencv to draw that by yourself, you need to save the image after the nvdsosd plugin.

edit_or · February 2, 2024, 6:02am

No. I still not succeed putting gst-dsexample after osd plug-in.
Any sample for that? I tried but didn’t work.
Thanks

yuweiw · February 2, 2024, 6:15am

Could you refer to this topic 279003? I have provided a source code for saving images with Bboxes. Please customize your own needs based on this.

edit_or · February 2, 2024, 6:17am

Ok thanks. Let me check. Thanks for the sample.

Topic		Replies	Views
Not able to draw line on incoming frame DeepStream SDK	17	2111	October 12, 2021
On JetsonNX, a segmentation fault occurred when using NvBufSurfTransform! DeepStream SDK	5	989	December 21, 2021
How to crop the image and save DeepStream SDK	47	15857	October 12, 2021
Plot bounding box on saved image DeepStream SDK	11	497	January 19, 2024
Segmentation fault when using nvds_obj_enc_process DeepStream SDK	9	490	April 26, 2024
Use the probe to take out the image error DeepStream SDK	6	579	March 10, 2023
Too Small Captured Full Frame to RAM in dsexample DeepStream SDK	8	869	October 12, 2022
Issue with plotting on image in deepstream-app DeepStream SDK	33	2185	October 12, 2021
Request for Assistance in Improving Output Render Quality DeepStream SDK jetson , deepstream	10	29	December 17, 2024
Segmentation fault in gst_dsexample.cpp DeepStream SDK	10	327	November 29, 2023

Segmentation fault on plotting rectangle on image

Related topics