I am using CSI camera on Beeagleboard ai-64. The model I am using is yolo5 from Ti Model Zoo.
I have done the configuration of CSI camera on Beagleboard. When testing with CSI camera works fine. For this I executed below command:
sudo gst-launch-1.0 v4l2src device=/dev/video2 \
! video/x-bayer, width=640, height=480, format=rggb \
! tiovxisp sink_0::device=/dev/v4l-subdev2 sensor-name=SENSOR_SONY_IMX219_RPI \
dcc-isp-file=/opt/imaging-REL.PSDK.JACINTO.09.00.00.02/sensor_drv/src/imx219/dcc_bins/dcc_viss_640x480.bin \
sink_0::dcc-2a-file=/opt/imaging-REL.PSDK.JACINTO.09.00.00.02/sensor_drv/src/imx219/dcc_bins/dcc_2a_640x480.bin \
format-msb=7 ! autovideosink
But my main work is to do object detection. I am executing below command for object detection:
sudo /opt/edge_ai_apps/apps_python/app_edgeai.py configs/object_detection.yaml
And it doesn’t give any output, but bunch of messages:
Number of subgraphs:1 , 129 nodes delegated out of 129 nodes
APP: Init ... !!!
MEM: Init ... !!!
MEM: Initialized DMA HEAP (fd=5) !!!
MEM: Init ... Done !!!
IPC: Init ... !!!
IPC: Init ... Done !!!
REMOTE_SERVICE: Init ... !!!
REMOTE_SERVICE: Init ... Done !!!
1372.717094 s: GTC Frequency = 200 MHz
APP: Init ... Done !!!
1372.717402 s: VX_ZONE_INIT:Enabled
1372.717425 s: VX_ZONE_ERROR:Enabled
1372.717430 s: VX_ZONE_WARNING:Enabled
1372.718230 s: VX_ZONE_INIT:[tivxInitLocal:130] Initialization Done !!!
1372.722929 s: VX_ZONE_INIT:[tivxHostInitLocal:86] Initialization Done for HOST !!!
[GST SRC STR]
[FLOW 0]
v4l2src device=/dev/video2 ! queue leaky=2 ! video/x-bayer, width=1920, height=1080, format=rggb ! tiovxisp sink_0::device=/dev/v4l-subdev2 sensor-name=SENSOR_SONY_IMX219_RPI dcc-isp-file=/opt/imaging/imx219/dcc_viss.bin sink_0::dcc-2a-file=/opt/imaging/imx219/dcc_2a.bin format-msb=7 ! video/x-raw, format=NV12 !tiovxmultiscaler name=split_01
split_01. ! queue ! video/x-raw, width=480, height=320 ! tiovxmultiscaler ! video/x-raw, width=320, height=320 ! tiovxdlpreproc data-type=3 target=0 channel-order=1 tensor-format=rgb out-pool-size=4 ! application/x-tensor-tiovx ! appsink name=pre_0 max-buffers=2 drop=true
split_01. ! queue ! video/x-raw, width=1280, height=720 ! tiovxdlcolorconvert target=1 out-pool-size=4 ! video/x-raw, format=RGB ! appsink name=sen_0 max-buffers=2 drop=true
[GST SINK STR]
appsrc format=GST_FORMAT_TIME is-live=true block=true do-timestamp=true name=post_0 ! tiovxdlcolorconvert ! video/x-raw,format=NV12, width=1280, height=720 ! queue ! mosaic_0.sink_0
tiovxmosaic name=mosaic_0 background=/tmp/background_0
sink_0::startx=320 sink_0::starty=180 sink_0::width=1280 sink_0::height=720
! video/x-raw,format=NV12, width=1920, height=1080 ! kmssink sync=false driver-name=tidss
1379.170143 s: VX_ZONE_INIT:[tivxHostDeInitLocal:100] De-Initialization Done for HOST !!!
1379.173713 s: VX_ZONE_INIT:[tivxDeInitLocal:193] De-Initialization Done !!!
APP: Deinit ... !!!
REMOTE_SERVICE: Deinit ... !!!
REMOTE_SERVICE: Deinit ... Done !!!
IPC: Deinit ... !!!
IPC: DeInit ... Done !!!
MEM: Deinit ... !!!
MEM: Alloc's: 28 alloc's of 70656496 bytes
MEM: Free's : 28 free's of 70656496 bytes
MEM: Open's : 0 allocs of 0 bytes
MEM: Deinit ... Done !!!
APP: Deinit ... Done !!!
PS: when I did the same for USB camera, the object detection was working.’
Kindly help if there’s any solution or am i doing something wrong.
Hi. I guess I can help you with RPi camera:
Please try this as for your RPi camera input in /opt/edge_ai_apps/config/image_classification.yaml:
input0:
source: /dev/video2
format: rggb
width: 1920
height: 1080
sen_id: imx219
framerate: 30
And you should check your /opt/edge_ai_apps/apps_python/gst_wrapper.py script for dcc-iso-file and sink_0 files paths.
Because in my case they point to /opt/imaging/%sen_id%/dcc_viss.bin and /opt/imaging/%sen_id%/dcc_2a.bin
So I had to take those files from: imaging.zip to /opt/imaging/imx219 folder.
But as I can see you already have them in /opt/imaging-REL.PSDK.JACINTO.09.00.00.02/sensor_drv/src/imx219/dcc_bins/
2 Likes
@Illia_Pikin
Thank you so much for your reply.
Well, I have two folders: Imaging and imaging-REL.PSDK.JACINTO.09.00.00.02.
But in gst_wrapper.py, the path is set for Imaging [ /opt/imaging]. Please see below image:
So am I doing something wrong?
1 Like
Hi. You can either edit the path in the gst_wrapper.py or simply copy /opt/imaging-REL.PSDK.JACINTO.09.00.00.02/sensor_drv/src/imx219/dcc_bins/dcc_viss.bin and /opt/imaging-REL.PSDK.JACINTO.09.00.00.02/sensor_drv/src/imx219/dcc_bins/dcc_2a.bin to /opt/imaging/imx219 folder.
sudo mkdir -p /opt/imaging/imx219
sudo cp /opt/imaging-REL.PSDK.JACINTO.09.00.00.02/sensor_drv/src/imx219/dcc_bins/dcc_viss.bin /opt/imaging-REL.PSDK.JACINTO.09.00.00.02/sensor_drv/src/imx219/dcc_bins/dcc_2a.bin /opt/imaging/imx219
2 Likes
i did the same.
I have set the path :
This is my folder:
After executing this command:
sudo /opt/edge_ai_apps/apps_python/app_edgeai.py configs/object_detection.yaml
The screen is stuck in this for few seconds
and outputs below logs:
debian@BeagleBone:/opt/edge_ai_apps$ sudo /opt/edge_ai_apps/apps_python/app_edgeai.py configs/object_detection.yaml
libtidl_onnxrt_EP loaded 0x2a3d6810
Final number of subgraphs created are : 1, - Offloaded Nodes - 391, Total Nodes - 391
APP: Init ... !!!
MEM: Init ... !!!
MEM: Initialized DMA HEAP (fd=4) !!!
MEM: Init ... Done !!!
IPC: Init ... !!!
IPC: Init ... Done !!!
REMOTE_SERVICE: Init ... !!!
REMOTE_SERVICE: Init ... Done !!!
22241.275562 s: GTC Frequency = 200 MHz
APP: Init ... Done !!!
22241.288885 s: VX_ZONE_INIT:Enabled
22241.288944 s: VX_ZONE_ERROR:Enabled
22241.288959 s: VX_ZONE_WARNING:Enabled
22241.308291 s: VX_ZONE_INIT:[tivxInitLocal:130] Initialization Done !!!
22241.311096 s: VX_ZONE_INIT:[tivxHostInitLocal:86] Initialization Done for HOST !!!
[GST SRC STR]
[FLOW 0]
v4l2src device=/dev/video2 ! queue leaky=2 ! video/x-bayer, width=640, height=480, format=rggb ! tiovxisp sink_0::device=/dev/v4l-subdev2 sensor-name=SENSOR_SONY_IMX219_RPI dcc-isp-file=/opt/imaging/src/imx219/dcc_viss.bin sink_0::dcc-2a-file=/opt/imaging/imx219/dcc_2a.bin format-msb=7 ! video/x-raw, format=NV12 !tiovxmultiscaler name=split_01
split_01. ! queue ! video/x-raw, width=640, height=480 ! videoscale ! video/x-raw, width=640, height=640 ! tiovxdlpreproc data-type=3 target=0 channel-order=0 tensor-format=rgb out-pool-size=4 ! application/x-tensor-tiovx ! appsink name=pre_0 max-buffers=2 drop=true
split_01. ! queue ! video/x-raw, width=640, height=480 ! tiovxdlcolorconvert target=1 out-pool-size=4 ! video/x-raw, format=RGB ! appsink name=sen_0 max-buffers=2 drop=true
[GST SINK STR]
appsrc format=GST_FORMAT_TIME is-live=true block=true do-timestamp=true name=post_0 ! tiovxdlcolorconvert ! video/x-raw,format=NV12, width=640, height=480 ! queue ! mosaic_0.sink_0
tiovxmosaic name=mosaic_0 background=/tmp/background_0
sink_0::startx=0 sink_0::starty=0 sink_0::width=640 sink_0::height=480
! video/x-raw,format=NV12, width=640, height=480 ! kmssink sync=false driver-name=tidss
22248.027693 s: VX_ZONE_INIT:[tivxHostDeInitLocal:100] De-Initialization Done for HOST !!!
22248.032311 s: VX_ZONE_INIT:[tivxDeInitLocal:193] De-Initialization Done !!!
APP: Deinit ... !!!
REMOTE_SERVICE: Deinit ... !!!
REMOTE_SERVICE: Deinit ... Done !!!
IPC: Deinit ... !!!
IPC: DeInit ... Done !!!
MEM: Deinit ... !!!
MEM: Alloc's: 25 alloc's of 49847828 bytes
MEM: Free's : 25 free's of 49847828 bytes
MEM: Open's : 0 allocs of 0 bytes
MEM: Deinit ... Done !!!
APP: Deinit ... Done !!!
Any help would be highly appreciated!
Try to use 1920 x 1080 camera resolution as input. It should be downscaled by gstreamer.
Or you can try to run this line (metioned here) in console before running your edgeai script:
media-ctl -d 0 --set-v4l2 '"imx219 6-0010":0[fmt:SRGGB8_1X8/640x480]'
thank you for your help.
I followed your steps and I got the below result.
I would like to know
- why there are 3 input videos and what does it mean?
- ALso it works with 1920x1080, but i want to do 640x480. it does not work for 640x480.
can you help?
1 Like
