I can run deepstream-test3 with IP Camera(e.g. ./deepstream-test3 rtsp://admin:1234@192.168.1.40:554/h264/ch1/main/av_stream).
But how to run deepstream-test3 with a usb camera, what kind of command line parameters should I input?
Can someone give me some advise? Thanks.
Hi,
You may get information about your usb cameras:
https://devtalk.nvidia.com/default/topic/1057681/jetson-tx1/logitech-c930e-on-jetson-tx1-very-slow-and-choppy-video/post/5363417/#5363417
And then replace uridecodebin with ‘v4l2src ! videoconvert ! nvvideoconvert’. Similar code is in
deepstream_sdk_v4.0_jetson\sources\apps\apps-common\src\deepstream_source_bin.c
It runs
v4l2src ! videoconvert ! nvvideoconvert ! 'video/x-raw(memory:NVMM),format=NV12'
In running deepstream-app, some settings are put in config file. For running deepstream-test*, you need to configure it:
v4l2src ! <b>video/x-raw,format=YUY2,width=848,height=480,framerate=30/1</b> ! videoconvert ! nvvideoconvert ! 'video/x-raw(memory:NVMM),format=NV12'
If you would like to run MJPEG source, you may refer to
https://devtalk.nvidia.com/default/topic/1058754/deepstream-sdk/how-to-change-source-pixel-format-from-yuyv-to-mjpg/post/5369809/#5369809
Hi, DaneLLL
Thanks for reply!
I have read the two posts you provided, but I still have two problems.
caps = gst_caps_new_simple ("video/x-raw", "format", G_TYPE_STRING, "YUY2",
"width", G_TYPE_INT, 848, "height", G_TYPE_INT,
480, "framerate", GST_TYPE_FRACTION,
30/1, config->source_fps_d, NULL);
Desperately need your help.
Thanks,
Newman97
Hi,
You should modify
deepstream_sdk_v4.0_jetson\sources\apps\sample_apps\deepstream-test3\deepstream_test3_app.c
deepstream_source_bin.c is for your reference.
Hi DaneLLL,
Thanks for your help, I create a function named “create_camera_source_bin” to replace “uridecodebin” as deepstream_source_bin.c does. But when I run it with command “./deepstream-test3-app v4l2src”(here the second parameter can be ignored), I encounter the following errors.
(deepstream-test3-app:15312): GStreamer-WARNING **: 20:06:48.863: Name 'src_cap_filter' is not unique in bin 'source-bin-00', not adding
Failed to link 'nvvidconv2, cap_filter'
Failed to create source bin. Exiting.
The error information shows link nvvidconv2 with cap_filter failed, but I have no idea about how to fix this problem, can you give me some advise?
Here is the code of “create_camera_source_bin” function I add to deepstream_test3_app.c.
static GstElement *
create_camera_source_bin (guint index, gchar * uri)
{
GstElement *bin = NULL;
GstCaps *caps = NULL;
gboolean ret = FALSE;
gchar bin_name[16] = { };
g_snprintf (bin_name, 15, "source-bin-%02d", index);
bin = gst_bin_new (bin_name);
GstElement *src_elem = gst_element_factory_make ("v4l2src", "src_elem");
if (!src_elem)
{
g_printerr ("Could not create 'src_elem'\n");
return NULL;
}
GstElement *cap_filter = gst_element_factory_make ("capsfilter", "src_cap_filter");
if (!cap_filter){
g_printerr ("Could not create 'src_cap_filter'\n");
return NULL;
}
caps = gst_caps_new_simple("video/x-raw", "format", G_TYPE_STRING, "YUY2",
"width", G_TYPE_INT, 848, "height", G_TYPE_INT, 480,
"framerate", GST_TYPE_FRACTION, 30, 1, NULL);
GstElement *nvvidconv1, *nvvidconv2;
GstCapsFeatures *feature = NULL;
nvvidconv1 = gst_element_factory_make ("videoconvert", "nvvidconv1");
if (!nvvidconv1){
g_printerr ("Failed to create 'nvvidcovn1'\n");
return NULL;
}
feature = gst_caps_features_new ("memory:NVMM", NULL);
gst_caps_set_features (caps, 0, feature);
g_object_set (G_OBJECT (cap_filter), "caps", caps, NULL);
nvvidconv2 = gst_element_factory_make ("nvvideoconvert", "nvvidconv2");
if (!nvvidconv2){
g_printerr ("Failed to create 'nvvidcovn2'\n");
return NULL;
}
g_object_set (G_OBJECT (nvvidconv2), "gpu-id", 0, NULL);
gst_bin_add_many (GST_BIN (bin), src_elem, cap_filter, nvvidconv1, nvvidconv2, cap_filter, NULL);
//NVGSTDS_LINK_ELEMENT (src_elem, nvvidconv1);
//NVGSTDS_LINK_ELEMENT (nvvidconv1, nvvidconv2);
//NVGSTDS_LINK_ELEMENT (nvvidconv2, cap_filter);
if (!gst_element_link (src_elem,nvvidconv1)){
g_printerr ("Failed to link 'src_elem, nvvidcovn1'\n");
return NULL;
}
if (!gst_element_link (nvvidconv1,nvvidconv2)){
g_printerr ("Failed to link 'nvvidcovn1, nvvidcovn2'\n");
return NULL;
}
if (!gst_element_link (nvvidconv2, cap_filter)){
g_printerr ("Failed to link 'nvvidcovn2, cap_filter'\n");
return NULL;
}
// NVGSTDS_BIN_ADD_GHOST_PAD (bin, cap_filter, "src");
GstPad *gstpad = gst_element_get_static_pad (cap_filter, "src");
gst_element_add_pad (bin, gst_ghost_pad_new ("src", gstpad));
gst_object_unref(gstpad);
gchar device[64];
g_snprintf (device, sizeof (device), "/dev/video%d", 0);
g_object_set (G_OBJECT (src_elem), "device", device, NULL);
return bin;
}
And I replace line 325 in deepstream_test3_app.c
GstElement *source_bin = create_source_bin (i, argv[i + 1]);
with
GstElement *source_bin = create_camera_source_bin (i, argv[i + 1]);
Thanks,
Newman97
Hi,
Do you run ‘v4l2-ctl’ command to get all modes? ‘848x480 YUYV 30fps’ is an example and you have to adapt it to your camera output.
Hi DaneLLL,
Yes I have run ‘v4l2-ctl’ command and hav tried different resolution and format config (my usb camera have ‘MJPG’ and ‘YUYV’ pixel format). Finally I solved the problem by replacing ‘YUY2’ format with ‘NV12’ format in line 25~27.
caps = gst_caps_new_simple("video/x-raw", "format", G_TYPE_STRING, "YUY2",
"width", G_TYPE_INT, 848, "height", G_TYPE_INT, 480,
"framerate", GST_TYPE_FRACTION, 30, 1, NULL);
with
caps = gst_caps_new_simple("video/x-raw", "format", G_TYPE_STRING, "NV12",
"width", G_TYPE_INT, 848, "height", G_TYPE_INT, 480,
"framerate", GST_TYPE_FRACTION, 30, 1, NULL);
It works now, but the video is slow and I am trying to solve it with the two reference post.
Thanks for your help.
Best regards,
Newman97
Hi,
You should create two cap_filter in the implementation. One is for v4l2src and the other is for nvvideoconvert.
For v4l2src, pick one supported mode:
video/x-raw,format=YUY2,width=848,height=480,framerate=30/1
For nvvideoconvert, it is
video/x-raw(memory:NVMM),format=NV12
It looks like you only create one ‘video/x-raw(memory:NVMM),format=YUY2,width=848,height=480,framerate=30/1’. It should not work.
Hi,
newman97
Could I get your code?
Thanks.
Hi DaneLLL,
I am trying to use USB camera on Jetson Nano. It runs smoothly with deepstream-app. I also tried using test-1 sample app with the code below but output is choppy. Could you please indicate what I am doing wrong?
Best regards.
static GstElement *
add_source_bin (void)
{
GstElement *src_bin = NULL;
gchar srcbin_name[16] = { };
GstCaps *caps = NULL;
GstCaps *lnvcaps = NULL;
src_bin = gst_bin_new (srcbin_name);
GstElement *src_elem = gst_element_factory_make (NVDS_ELEM_SRC_CAMERA_V4L2, "src_elem");
gchar device[64];
g_snprintf (device, sizeof (device), "/dev/video%d", 0);
g_object_set (G_OBJECT (src_elem), "device", device, NULL);
GstElement *cap_filter = gst_element_factory_make (NVDS_ELEM_CAPS_FILTER, "src_cap_filter");
GstElement *nvvidconv1, *nvvidconv2;
GstCapsFeatures *feature = NULL;
nvvidconv1 = gst_element_factory_make ("videoconvert", "nvvidconv1");
nvvidconv2 = gst_element_factory_make (NVDS_ELEM_VIDEO_CONV, "nvvidconv2");
g_object_set (G_OBJECT (nvvidconv2), "gpu-id", 0, "nvbuf-memory-type", 0, NULL);
caps = gst_caps_new_simple("video/x-raw","format", G_TYPE_STRING, "YUY2",
"width", G_TYPE_INT, 800, "height", G_TYPE_INT, 600,
"framerate", GST_TYPE_FRACTION, 30,1, NULL);
lnvcaps = gst_caps_new_simple("video/x-raw", "format", G_TYPE_STRING,"NV12",
"width", G_TYPE_INT, 480, "height", G_TYPE_INT, 272,
"framerate", GST_TYPE_FRACTION, 30,1, NULL);
feature = gst_caps_features_new ("memory:NVMM", NULL);
gst_caps_set_features (lnvcaps, 0, feature);
g_object_set (G_OBJECT (cap_filter), "caps", lnvcaps, NULL);
gst_bin_add_many (GST_BIN (src_bin), src_elem, nvvidconv1, nvvidconv2, cap_filter, NULL);
if (!gst_element_link_filtered (src_elem, nvvidconv1, caps)){
g_printerr ("Failed to link 'src_elem, nvvidconv1'\n");
return NULL;
}
if (!gst_element_link (nvvidconv1,nvvidconv2)){
g_printerr ("Failed to link 'nvvidconv1, nvvidconv2'\n");
return NULL;
}
if (!gst_element_link (nvvidconv2, cap_filter)){
g_printerr ("Failed to link 'nvvidconv2, cap_filter'\n");
return NULL;
}
GstPad *gstpad = gst_element_get_static_pad (cap_filter, "src");
gst_element_add_pad (src_bin, gst_ghost_pad_new ("src", gstpad));
gst_object_unref(gstpad);
gst_caps_unref (caps);
gst_caps_unref (lnvcaps);
return src_bin;
}
int
main (int argc, char *argv[])
{
GMainLoop *loop = NULL;
GstElement *pipeline = NULL, *streammux = NULL, *sink = NULL, *pgie = NULL, *nvvidconv = NULL,
*nvosd = NULL;
GstElement *transform = NULL;
GstBus *bus = NULL;
guint bus_watch_id;
GstPad *osd_sink_pad = NULL;
/* Standard GStreamer initialization */
gst_init (&argc, &argv);
loop = g_main_loop_new (NULL, FALSE);
/* Create gstreamer elements */
pipeline = gst_pipeline_new ("dstest1-pipeline");
GstElement *source = add_source_bin();
streammux = gst_element_factory_make ("nvstreammux", "stream-muxer");
pgie = gst_element_factory_make ("nvinfer", "primary-nvinference-engine");
nvvidconv = gst_element_factory_make ("nvvideoconvert", "nvvideo-converter");
g_object_set (G_OBJECT (nvvidconv), "gpu-id", 0, NULL);
nvosd = gst_element_factory_make ("nvdsosd", "nv-onscreendisplay");
transform = gst_element_factory_make ("nvegltransform", "nvegl-transform");
sink = gst_element_factory_make ("nveglglessink", "nvvideo-renderer");
g_object_set (G_OBJECT (streammux), "width", MUXER_OUTPUT_WIDTH, "height",
MUXER_OUTPUT_HEIGHT, "batch-size", 1,
"batched-push-timeout", MUXER_BATCH_TIMEOUT_USEC, NULL);
g_object_set (G_OBJECT (pgie),
"config-file-path", "dstest1_pgie_config.txt", NULL);
/* we add a message handler */
bus = gst_pipeline_get_bus (GST_PIPELINE (pipeline));
bus_watch_id = gst_bus_add_watch (bus, bus_call, loop);
gst_object_unref (bus);
/* Set up the pipeline */
gst_bin_add_many (GST_BIN (pipeline),
source, streammux, pgie, nvvidconv, nvosd, transform, sink, NULL);
GstPad *sinkpad, *srcpad;
gchar pad_name_sink[16] = "sink_0";
gchar pad_name_src[16] = "src";
sinkpad = gst_element_get_request_pad (streammux, pad_name_sink);
if (!sinkpad) {
g_printerr ("Streammux request sink pad failed. Exiting.\n");
return -1;
}
srcpad = gst_element_get_static_pad (source, pad_name_src);
if (!srcpad) {
g_printerr ("Source request src pad failed. Exiting.\n");
return -1;
}
if (gst_pad_link (srcpad, sinkpad) != GST_PAD_LINK_OK) {
g_printerr ("Failed to link decoder to stream muxer. Exiting.\n");
return -1;
}
gst_object_unref (sinkpad);
gst_object_unref (srcpad);
/* we link the elements together */
if (!gst_element_link_many (streammux, pgie,
nvvidconv, nvosd, transform, sink, NULL)) {
g_printerr ("Elements could not be linked: 2. Exiting.\n");
return -1;
}
/* Set the pipeline to "playing" state */
gst_element_set_state (pipeline, GST_STATE_PLAYING);
/* Wait till pipeline encounters an error or EOS */
g_main_loop_run (loop);
/* Out of the main loop, clean up nicely */
g_print ("Returned, stopping playback\n");
gst_element_set_state (pipeline, GST_STATE_NULL);
g_print ("Deleting pipeline\n");
gst_object_unref (GST_OBJECT (pipeline));
g_source_remove (bus_watch_id);
g_main_loop_unref (loop);
return 0;
}
Now it works smoothly.
Thanks a lot
Hi @DaneLLL,
I am having trouble applying that patch to this code. Things are different enough that I am lost.
Here is the function that works for me for two cameras, but it is slow as expected.
static GstElement *
create_camera_source_bin (guint index, gchar * uri)
{
GstElement *bin = NULL;
GstCaps *caps = NULL;
gboolean ret = FALSE;
gchar bin_name[16] = { };
g_snprintf (bin_name, 15, "source-bin-%02d", index);
bin = gst_bin_new (bin_name);
GstElement *src_elem = gst_element_factory_make ("v4l2src", "src_elem");
if (!src_elem)
{
g_printerr ("Could not create 'src_elem'\n");
return NULL;
}
gchar cap_filt_char[64];
g_snprintf (cap_filt_char, sizeof (cap_filt_char), "src_cap_filter_%d", index);
GstElement *cap_filter = gst_element_factory_make ("capsfilter", cap_filt_char);
if (!cap_filter){
g_printerr ("Could not create 'src_cap_filter'\n");
return NULL;
}
caps = gst_caps_new_simple("video/x-raw", "format", G_TYPE_STRING, "NV12",
"width", G_TYPE_INT, 1920, "height", G_TYPE_INT, 1080,
"framerate", GST_TYPE_FRACTION, 60, 1, NULL);
GstElement *nvvidconv1, *nvvidconv2;
GstCapsFeatures *feature = NULL;
nvvidconv1 = gst_element_factory_make ("videoconvert", "nvvidconv1");
if (!nvvidconv1){
g_printerr ("Failed to create 'nvvidcovn1'\n");
return NULL;
}
feature = gst_caps_features_new ("memory:NVMM", NULL);
gst_caps_set_features (caps, 0, feature);
g_object_set (G_OBJECT (cap_filter), "caps", caps, NULL);
nvvidconv2 = gst_element_factory_make ("nvvideoconvert", "nvvidconv2");
if (!nvvidconv2){
g_printerr ("Failed to create 'nvvidcovn2'\n");
return NULL;
}
g_object_set (G_OBJECT (nvvidconv2), "gpu-id", 0, NULL);
gst_bin_add_many (GST_BIN (bin), src_elem, cap_filter, nvvidconv1, nvvidconv2, cap_filter, NULL);
//NVGSTDS_LINK_ELEMENT (src_elem, nvvidconv1);
//NVGSTDS_LINK_ELEMENT (nvvidconv1, nvvidconv2);
//NVGSTDS_LINK_ELEMENT (nvvidconv2, cap_filter);
if (!gst_element_link (src_elem,nvvidconv1)){
g_printerr ("Failed to link 'src_elem, nvvidcovn1'\n");
return NULL;
}
if (!gst_element_link (nvvidconv1,nvvidconv2)){
g_printerr ("Failed to link 'nvvidcovn1, nvvidcovn2'\n");
return NULL;
}
if (!gst_element_link (nvvidconv2, cap_filter)){
g_printerr ("Failed to link 'nvvidcovn2, cap_filter'\n");
return NULL;
}
// NVGSTDS_BIN_ADD_GHOST_PAD (bin, cap_filter, "src");
GstPad *gstpad = gst_element_get_static_pad (cap_filter, "src");
gst_element_add_pad (bin, gst_ghost_pad_new ("src", gstpad));
gst_object_unref(gstpad);
gchar device[64];
g_snprintf (device, sizeof (device), "/dev/video%d", index);
g_object_set (G_OBJECT (src_elem), "device", device, NULL);
return bin;
}
How would I apply that patch to this example? I find all this gstreamer code to be pretty non-intuitive. My cameras are Econsystems USB3 cameras with UYVY output. I am using DS5.0 and the Jetson Nano.
Thank you!
And here is my attempt to merge in the patch. It does not work, throwing the error “streaming stopped, reason not-linked”.
static GstElement *
create_camera_source_bin (guint index, gchar * uri)
{
GstElement *bin = NULL;
GstCaps *caps = NULL;
GstCapsFeatures *feature = NULL;
gboolean ret = FALSE;
// GstElement *nvvidconv1, *nvvidconv2;
GstElement *conv1 = NULL, *conv2 = NULL, *nvconv = NULL;
GstCaps *caps_uyvy = NULL, *caps_nv12_nvmm = NULL, *caps_nv12 = NULL;
gchar bin_name[16] = { };
g_snprintf (bin_name, 15, "source-bin-%02d", index);
bin = gst_bin_new (bin_name);
GstElement *src_elem = gst_element_factory_make ("v4l2src", "src_elem");
gchar cap_filt_char[64];
g_snprintf (cap_filt_char, sizeof (cap_filt_char), "src_cap_filter_%d", index);
GstElement *cap_filter = gst_element_factory_make ("capsfilter", cap_filt_char);
conv1 = gst_element_factory_make ("nvvidconv", "conv1");
conv2 = gst_element_factory_make ("nvvidconv", "conv2");
caps_uyvy = gst_caps_new_simple ("video/x-raw", "format", G_TYPE_STRING, "UYVY",
"width", G_TYPE_INT, 1920, "height", G_TYPE_INT,
1080, "framerate", GST_TYPE_FRACTION,
60, 1, NULL);
caps_nv12_nvmm = gst_caps_new_simple ("video/x-raw", "format", G_TYPE_STRING, "NV12", NULL);
feature = gst_caps_features_new ("memory:NVMM", NULL);
gst_caps_set_features (caps_nv12_nvmm, 0, feature);
g_object_set (G_OBJECT (cap_filter), "caps", caps_nv12_nvmm, NULL);
caps_nv12 = gst_caps_new_simple ("video/x-raw", "format", G_TYPE_STRING, "NV12", NULL);
//nvvidconv2 = gst_element_factory_make ("nvvideoconvert", "nvvidconv2");
nvconv = gst_element_factory_make ("nvvideoconvert", "nvconv");
gchar device[64];
g_snprintf (device, sizeof (device), "/dev/video%d", index);
g_object_set (G_OBJECT (src_elem), "device", device, NULL);
// g_object_set (G_OBJECT (nvvidconv2), "gpu-id", 0, NULL);
gst_bin_add_many (GST_BIN (bin), src_elem, cap_filter, conv1, nvconv, NULL);
gst_element_link_filtered(src_elem, conv1, caps_uyvy);
gst_bin_add (GST_BIN (bin), conv2);
gst_element_link_filtered(conv1, conv2, caps_nv12_nvmm);
gst_element_link_filtered(conv2, nvconv, caps_nv12);
gst_element_link_filtered(nvconv, cap_filter, caps_nv12_nvmm);
// NVGSTDS_BIN_ADD_GHOST_PAD (bin, cap_filter, "src");
GstPad *gstpad = gst_element_get_static_pad (cap_filter, "src");
gst_element_add_pad (bin, gst_ghost_pad_new ("src", gstpad));
gst_object_unref(gstpad);
return bin;
}
This code is merged in to deepstream_test3_app.c from DS5.0. As mentioned above you must also replace line 325 in deepstream_test3_app.c
GstElement *source_bin = create_source_bin (i, argv[i + 1]);
with
GstElement *source_bin = create_camera_source_bin (i, argv[i + 1]);
Any help is appreciated!
Hi,
By default it is uridecodebin in deepstream-test3. You need to replace it with
v4l2src ! video/x-raw,format=_SOURCE_FORMAT_,width=_SOURCE_WIDTH_,height=SOURCE_HEIGHT_,framerate=_SOURCE_FRAMERATR_ ! videoconvert ! nvvideoconvert ! video/x-raw(memory:NVMM),format=NV12
You can refer to this post to set precise format,width,height,framerate.
Thank you, but I don’t understand how to convert these pipeline commands in to C++ code. I wasn’t able to understand it from reading the code. I suppose I will have to find a tutorial that explains it. Do you know of such a tutorial?
I did succeed in getting fast video processing with one camera using the recommended modifications to deepstream-test-1, I just do not get how to update deepstream-test-3 to do the same with multiple cameras.
Is there anyway to skip videoconvert plugin?
Hi,
We will have enhancement in next release. Please wait for DS 5.0 GA.