# Error generated while running the code after connecting the camera **URL:** \ **Category:** Jetson Xavier NX\ **Tags:** nvbugs, gstreamer\ **Created:** [November 15, 2023, 12:52pm UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958 "2023-11-15T12:52:17Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 15, 2023, 12:52pm UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/1 "2023-11-15T12:52:17Z")
Continuing the discussion from [Nano not using GPU with gstreamer/python. Slow FPS, dropped frames](https://forums.developer.nvidia.com/t/nano-not-using-gpu-with-gstreamer-python-slow-fps-dropped-frames/107291/22): ![image](https://global.discourse-cdn.com/nvidia/original/4X/3/2/9/329278e33ec224dd4cf9b6e3805661a8685b382c.png) ---
**Author:** ![DaneLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/danelll/32/14043_2.png) [@DaneLLL](https://forums.developer.nvidia.com/u/DaneLLL)\ **Post date:** [November 16, 2023, 2:07am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/3 "2023-11-16T02:07:28Z")
Hi, Please check if CUDA\_VER is correctly set. You can run _$ ls /usr/local/_ to get the version. ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 16, 2023, 5:09am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/4 "2023-11-16T05:09:37Z")
Dear DaneLL, I have verified the cuda version ,In my case i am using 11.4 , [gst\_cv\_gpumat.zip](https://forums.developer.nvidia.com/uploads/short-url/nKnb2H1Uw6Z80eFP2WS6OPUtSaf.zip) (3.0 KB) I have no problem during the make ,but when i run executable it is showing this error . CUDA : 11.4 Platform: Jetson xavier NX Thanks in advance ---
**Author:** ![DaneLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/danelll/32/14043_2.png) [@DaneLLL](https://forums.developer.nvidia.com/u/DaneLLL)\ **Post date:** [November 16, 2023, 5:37am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/5 "2023-11-16T05:37:50Z")
Hi, The sample is for Jetpack 4 release and does not work on Jetpack 5, since Nvbuffer APIs are deprecated on Jetpack 5. You would need to replace the Nvbuffer APIs with NvBufSurface APIs. ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 16, 2023, 5:53am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/6 "2023-11-16T05:53:51Z")
Thanks for the quick update , is it like this ![image](https://global.discourse-cdn.com/nvidia/original/4X/a/c/1/ac165d9f595de598365f54978fdc6ef6eb2e918c.png) ,But i got the error… @DaneLLL ---
**Author:** ![DaneLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/danelll/32/14043_2.png) [@DaneLLL](https://forums.developer.nvidia.com/u/DaneLLL)\ **Post date:** [November 16, 2023, 7:04am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/7 "2023-11-16T07:04:23Z")
Hi, You may refer to this patch to use NvBufurface APIs: [How to create opencv gpumat from nvstream? - #18 by DaneLLL](https://forums.developer.nvidia.com/t/how-to-create-opencv-gpumat-from-nvstream/70680/18) We will also check and try to have a sample for Jetpack 5. Not sure which version you are using. It would be great if you can use latest Jetpack 5.1.2. ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 16, 2023, 9:13am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/10 "2023-11-16T09:13:45Z")
Thanks Dane, ```auto #include #include #include #include #include "nvbuf_utils.h" #include #include #include #include #include using namespace std; #define USE(x) ((void)(x)) static GstPipeline *gst_pipeline = nullptr; static string launch_string; GstClockTime usec = 1000000; static int w = 1920; static int h = 1080; EGLDisplay egl_display; static bool create_filter = true; static cv::Ptr< cv::cuda::Filter > filter; static GstPadProbeReturn conv_src_pad_buffer_probe (GstPad * pad, GstPadProbeInfo * info, gpointer u_data) { GstBuffer *buffer = (GstBuffer *) info->data; GstMapInfo map = {0}; int dmabuf_fd = 0; CUresult status; gst_buffer_map (buffer, &map, GST_MAP_READ); ExtractFdFromNvBuffer((void *)map.data, &dmabuf_fd); //CUDA postprocess { EGLImageKHR egl_image; egl_image = NvEGLImageFromFd(egl_display, dmabuf_fd); CUresult status; CUeglFrame eglFrame; CUgraphicsResource pResource = NULL; cudaFree(0); // status = cuGraphicsEGLRegisterImage(&pResource, // egl_image, // CU_GRAPHICS_MAP_RESOURCE_FLAGS_NONE); // if (status != CUDA_SUCCESS) // { // printf("cuGraphicsEGLRegisterImage failed: %d \n",status); // } // status = cuGraphicsResourceGetMappedEglFrame(&eglFrame, pResource, 0, 0); // status = cuCtxSynchronize(); status = cuGraphicsEGLRegisterImage(&pResource, egl_image, CU_GRAPHICS_MAP_RESOURCE_FLAGS_NONE); if (status != CUDA_SUCCESS) { printf("cuGraphicsEGLRegisterImage failed: %d \n",status); } status = cuGraphicsResourceGetMappedEglFrame(&eglFrame, pResource, 0, 0); status = cuCtxSynchronize(); if (create_filter) { filter = cv::cuda::createSobelFilter(CV_8UC4, CV_8UC4, 1, 0, 3, 1, cv::BORDER_DEFAULT); //filter = cv::cuda::createGaussianFilter(CV_8UC4, CV_8UC4, cv::Size(31,31), 0, 0, cv::BORDER_DEFAULT); create_filter = false; } cv::cuda::GpuMat d_mat(h, w, CV_8UC4, eglFrame.frame.pPitch[0]); filter->apply (d_mat, d_mat); status = cuCtxSynchronize(); status = cuGraphicsUnregisterResource(pResource); NvDestroyEGLImage(egl_display, egl_image); } gst_buffer_unmap(buffer, &map); return GST_PAD_PROBE_OK; } int main(int argc, char** argv) { USE(argc); USE(argv); gst_init (&argc, &argv); GMainLoop *main_loop; main_loop = g_main_loop_new (NULL, FALSE); ostringstream launch_stream; egl_display = eglGetDisplay(EGL_DEFAULT_DISPLAY); eglInitialize(egl_display, NULL, NULL); launch_stream << "nvarguscamerasrc sensor-id=0 ! " << "video/x-raw(memory:NVMM),width="<< w <<",height="<< h <<",framerate=30/1,format=NV12 ! " << "nvvidconv name=myconv ! " << "video/x-raw(memory:NVMM),format=RGBA ! " << "nvoverlaysink "; launch_string = launch_stream.str(); g_print("Using launch string: %s\n", launch_string.c_str()); GError *error = nullptr; gst_pipeline = (GstPipeline*) gst_parse_launch(launch_string.c_str(), &error); if (gst_pipeline == nullptr) { g_print( "Failed to parse launch: %s\n", error->message); return -1; } if(error) g_error_free(error); GstElement* conv = gst_bin_get_by_name(GST_BIN(gst_pipeline), "myconv"); GstPad *src_pad = gst_element_get_static_pad (conv, "src"); gst_pad_add_probe (src_pad, GST_PAD_PROBE_TYPE_BUFFER, conv_src_pad_buffer_probe, NULL, NULL); gst_element_set_state((GstElement*)gst_pipeline, GST_STATE_PLAYING); g_usleep(30*usec); GstElement* src = gst_bin_get_by_name(GST_BIN(gst_pipeline), "mysource"); gst_element_send_event (src, gst_event_new_eos ()); // Wait for EOS message GstBus *bus = gst_pipeline_get_bus(GST_PIPELINE(gst_pipeline)); gst_bus_poll(bus, GST_MESSAGE_EOS, GST_CLOCK_TIME_NONE); gst_element_set_state((GstElement*)gst_pipeline, GST_STATE_NULL); gst_object_unref(GST_OBJECT(gst_pipeline)); g_main_loop_unref(main_loop); eglTerminate(egl_display); g_print("going to exit \n"); return 0; } ``` I have edited the code but still i am facing an error during the run time ![image](https://global.discourse-cdn.com/nvidia/original/4X/7/6/6/766c9c37b2d541253a1f095e7f4cc3d8924ea23d.png) **Hi,One more thing i am not using Deepstream ,the given code is extracting the frame captured using nvarguscamera and place that to the cv::cuda::GPuMat from NVMM .** ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 18, 2023, 6:23am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/11 "2023-11-18T06:23:24Z")
@DaneLLL Sir can you please give me some input…Thanks in advance ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 20, 2023, 5:21am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/12 "2023-11-20T05:21:40Z")
@DaneLLL @DaveYYY Guys can you please give a support thanks in advance… ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 20, 2023, 6:24am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/14 "2023-11-20T06:24:11Z")
Thanks verymuch , But when i tried to run it is not showing any output. @DaveYYY Thanks in advance ---
**Author:** ![DaveYYY](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/daveyyy/32/233244_2.png) [@DaveYYY](https://forums.developer.nvidia.com/u/DaveYYY)\ **Post date:** [November 20, 2023, 7:00am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/15 "2023-11-20T07:00:37Z")
Hi, Sorry that I missed one thing here. `nvoverlaysink` is deprecated on JetPack 5, and please use `nv3dsink` instead. ```auto launch_stream << "nvarguscamerasrc name=mysource ! " << "video/x-raw(memory:NVMM),width="<< w <<",height="<< h <<",framerate=30/1,format=NV12 ! " << "nvvidconv name=myconv ! " << "video/x-raw(memory:NVMM),format=RGBA ! " << "nv3dsink "; ``` [gst\_cv\_gpumat.tbz2](https://forums.developer.nvidia.com/uploads/short-url/ow56rHrWd27YvgKxMohjklPVjLf.tbz2) (3.4 KB) ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 20, 2023, 8:53am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/16 "2023-11-20T08:53:58Z")
THANKZZZZZZZZZZZZZZZZZZZZZ… you are an ultra saver… I am trying to capture the camera in 60FPS so i had made a tweak in the code ```auto << "nvarguscamerasrc name=mysource ! " << "video/x-raw(memory:NVMM),width="<< w <<",height="<< h <<",framerate=60/1,format=NV12 ! " << "nvvidconv name=myconv ! " << "video/x-raw(memory:NVMM),format=RGBA ! " << "nv3dsink "; ``` But ends up in error ,Can you please give me some input on that …Thanks in advance @DaveYYY ---
**Author:** ![DaneLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/danelll/32/14043_2.png) [@DaneLLL](https://forums.developer.nvidia.com/u/DaneLLL)\ **Post date:** [November 20, 2023, 8:57am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/18 "2023-11-20T08:57:40Z")
Hi, Please run your pipeline in _gst-launch-1.0_ to make sure it is valid first. And then apply to C code. ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 24, 2023, 3:45am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/19 "2023-11-24T03:45:34Z")
I had done all the suggestions from your side and it succeeded well. Some more questions 1,I _ **triying to run the NVMM code in one thread and needs to place the GpuMat variable to a Queue so i can acess the queue from another thread.** _ _How can i place the GpuMat variable to a normal queue ,Is it possible or Is there any method to place the GpuMat variable to a queue in GPU?_ 1. _Which is faster to process ,in my case i am using 1080p RGB video ,Is it queue or shared memory or **any other faster method is available**?_ 3._I want to blend an rgba image over this Gpu Mat image .From the documention i am able to see that using **VIC(Video image compositor)** we can able to blend on images .Can you please help me that how can i blend my RGBA image to normal image using the VIC engine or any sample code available?_ **Thanks in advance…** @DaveYYY @DaneLLL ---
**Author:** ![DaneLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/danelll/32/14043_2.png) [@DaneLLL](https://forums.developer.nvidia.com/u/DaneLLL)\ **Post date:** [November 24, 2023, 5:23am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/20 "2023-11-24T05:23:52Z")
Hi, > [@manuel.johnson](#): > > How can i place the GpuMat variable to a normal queue ,Is it possible or Is there any method to place the GpuMat variable to a queue in GPU? The demonstration is to map NvBufSurface to a GpuMat. So a possible solution should be pass NvBufSurface to another thread, and map it to GpuMat. Please try this. Or see if other users can share experience in OpenCV programming. > [@manuel.johnson](#): > > Which is faster to process ,in my case i am using 1080p RGB video ,Is it queue or shared memory or **any other faster method is available**? There may be no existing implementation for this. Probably you can implement the function in CUDA code so that it can utilize GPU engine > [@manuel.johnson](#): > > _I want to blend an rgba image over this Gpu Mat image .From the documention i am able to see that using **VIC(Video image compositor)** we can able to blend on images .Can you please help me that how can i blend my RGBA image to normal image using the VIC engine or any sample code available?_ You can try _NvBufSurfTransformMultiInputBufCompositeBlend()_. There is demonstration in > /usr/src/jetson\_multimedia\_api/samples/13\_argus\_multi\_camera Also can refer to source code of _nvcompositor_ plugin. ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 24, 2023, 5:55am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/21 "2023-11-24T05:55:11Z")
Thanks for the quick update , I have a rgba image .I need to blend that image onto a video .(Consider the background as a video and dice as an rgba image ) Like below ![8KvHk](https://global.discourse-cdn.com/nvidia/original/4X/2/d/b/2dbe4e552a6e610010c30866e80454869384003a.jpeg) is that possible with the mention sample app …Dane … Thanks in advance… @DaveYYY @DaneLLL ---
**Author:** ![DaneLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/danelll/32/14043_2.png) [@DaneLLL](https://forums.developer.nvidia.com/u/DaneLLL)\ **Post date:** [November 24, 2023, 6:21am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/22 "2023-11-24T06:21:08Z")
Hi, Transparent pixels are not supported. Please refer to discussion in the topic: [Nvcompositor plugs alpha is not work and alpha plugs no work](https://forums.developer.nvidia.com/t/nvcompositor-plugs-alpha-is-not-work-and-alpha-plugs-no-work/230557) The whole image can be blended in single alpha value. For this use-case the dices have alpha=1 and other pixels have alpha=0(transparent). This is not supported in hardware converter. ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 24, 2023, 6:34am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/23 "2023-11-24T06:34:15Z")
Many Many thanks , Can you please give me any other faster method . ```auto cv::Mat overlayImages_cuda(const cv::cuda::GpuMat& face_img) { cv::cuda::GpuMat resized_image_torric; // cv::resize(image_torric, resized_image_torric, face_img.size(), 0, 0, cv::INTER_NEAREST); cv::cuda::resize(transImg, resized_image_torric, face_img.size(), 0, 0, cv::INTER_NEAREST); std::vector channels; cv::cuda::GpuMat overlay_img,result; cv::cuda::GpuMat overlay_mask,background_mask,face_part,overlay_part; cv::cuda::split(resized_image_torric, channels); cv::cuda::merge({channels[0], channels[1], channels[2]}, overlay_img); channels[3].copyTo(overlay_mask); cv::cuda::subtract(cv::Scalar(255), overlay_mask, background_mask); face_img.convertTo(face_part, CV_32F,1.0 / 255.0); overlay_img.convertTo(overlay_part, CV_32F, 1.0 / 255.0); background_mask.convertTo(background_mask, CV_32F,1.0 / 255.0); overlay_mask.convertTo(overlay_mask, CV_32F, 1.0 / 255.0); cv::cuda::cvtColor(background_mask, background_mask, cv::COLOR_GRAY2BGR); cv::cuda::cvtColor(overlay_mask,overlay_mask, cv::COLOR_GRAY2BGR); cv::cuda::multiply(face_part, background_mask,face_part); cv::cuda::multiply(overlay_part, overlay_mask, overlay_part); cv::cuda::addWeighted(face_part, 255.0, overlay_part, 255.0, 0.0, result); result.convertTo(result, CV_8U); //Download from the gpu image part------------------------------------------------------ cv::Mat temp_frame_downloaded; result.download(temp_frame_downloaded); // cv::imshow("Object Detection", temp_frame_downloaded); return temp_frame_downloaded; } ``` I had tried this method but it is taking around 50ms for the execution of a single frame (1920x1080 @30fps) i want to reduce it to around 10ms ,Can you please give me some suggestion,How can i blend it. @DaveYYY @DaneLLL ---
**Author:** ![DaneLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/danelll/32/14043_2.png) [@DaneLLL](https://forums.developer.nvidia.com/u/DaneLLL)\ **Post date:** [November 24, 2023, 7:22am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/24 "2023-11-24T07:22:53Z")
Hi, We don’t know if there is other existing solution for the use-case. Would need other users to share experience. One more thing you may try is to run _sudo jetson\_clocks_. This enables GPU engine at maximum clock. Please try and see if it brings enhancement. ---
**Author:** ![manuel.johnson](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/manuel.johnson/32/267115_2.png) [@manuel.johnson](https://forums.developer.nvidia.com/u/manuel.johnson)\ **Post date:** [November 28, 2023, 3:10am UTC](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958/25 "2023-11-28T03:10:10Z")
Thanks for your kind response , As per your suggestion maximise the clock perfomance i have tried out.But not much difference Can you suggest any method other than cv2.cuda.addweighted api .In nvidia (Blending RGBA image with Normal image) I have gone through a lot of articles but things got in to vain… Your suggestions are really appreciated… @DaveYYY @DaneLLL [Next page](https://forums.developer.nvidia.com/t/error-generated-while-running-the-code-after-connecting-the-camera/272958.md?page=2)