Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson)
• DeepStream Version 6.3
• JetPack Version (valid for Jetson only) 5.1.2
How to correctly use the roi-filtering function in the Gst-nvdsanalytics plugin in deepstream-yolo and transfer the result data of the detection to the outside?
First question
I’m using the Gst-nvdsanalytics feature in deepstream and I’ve successfully implemented it Overcrowding Detection, Direction Detection, Line Crossing, but it seems that roi-filtering doesn’t work, why is this?
Could it be that there is a problem with my configuration parameters? I’ve tried it many times and it doesn’t show up in the output, why is this?
This is my config_nvdsanalytics, configuration parameters, and I set enable=1, but it doesn’t seem to work.
[property]
enable=1
config-width=1920
config-height=1080
osd-mode=2
display-font-size=12
[roi-filtering-stream-0]
#enable or disable following feature
enable=1
#ROI to filter select objects, and remove from meta data
roi-RF=444;357;995;357;1225;424;1860;882;4;882;
#remove objects in the ROI
inverse-roi=0
class-id=-1
[overcrowding-stream-0]
enable=0
roi-OC=444;357;995;357;1225;424;1860;882;4;882;
object-threshold=15
class-id=-1
[line-crossing-stream-0]
enable=0
line-crossing-Entry=566;611;428;914;158;738;854;732
line-crossing-Exit=1407;880;1227;620;895;732;1672;732
class-id=-1
extended=0
mode=loose
[direction-detection-stream-0]
enable=0
#Label;direction;
direction-South=566;611;428;914;
direction-North=1407;880;1227;620;
class-id=-1
This is my deepstream_app_config parameter configuration! I guess they should be fine?
[application]
enable-perf-measurement=1
perf-measurement-interval-sec=1
[tiled-display]
enable=1
rows=1
columns=1
width=1280
height=720
gpu-id=0
nvbuf-memory-type=0
[source0]
enable=1
type=2
uri=file:///home/cvai/lintao/ultralytics/main/video/LT_cloudy_morning_sub_output.mp4
#uri=file:///home/cvai/lintao/ultralytics/main/video/traffic.mp4
num-sources=1
num-extra-surfaces=10
drop-frame-interval=0
nvbuf-memory-type=0
gpu-id=0
[sink0]
enable=1
type=2
sync=0
gpu-id=0
nvbuf-memory-type=0
[tracker]
enable=1
tracker-width=640
tracker-height=640
ll-lib-file=/opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so
ll-config-file=/home/cvai/lintao/ultralytics/DeepStream-Yolo-master/config_tracker_NvDCF_perf.yml
gpu-id=0
display-tracking-id=1
user-meta-pool-size=256
[osd]
enable=1
gpu-id=0
border-width=5
text-size=15
text-color=1;1;1;1;
text-bg-color=0.3;0.3;0.3;1
font=Serif
show-clock=1
clock-x-offset=1000
clock-y-offset=30
clock-text-size=15
clock-color=1;0;0;0
nvbuf-memory-type=0
[streammux]
gpu-id=0
live-source=0
batch-size=1
batched-push-timeout=40000
width=1280
height=720
enable-padding=0
nvbuf-memory-type=0
[primary-gie]
enable=1
gpu-id=0
gie-unique-id=1
nvbuf-memory-type=0
config-file=config_infer_primary_yoloV8.txt
[tests]
file-loop=0
[nvds-analytics]
enable=1
config-file=config_nvdsanalytics.txt
But the reality is that when I only set enable=1 in [roi-filtering-stream-0], there is no output, and even the ROI area is not displayed, as shown below:
If I set [overcrowding-stream-0] and [line-crossing-stream-0] to enable=1, it works!
So how should I achieve this? I refer to the example in the example as well as the official docs, my setup should be fine, but it doesn’t work?
If there is a problem, can you tell me how it should be set up correctly?
Second question
Can I get the detection data in it,
For example, the detection data of crowd detection in the figure above, as well as the data obtained by cross-line counting.
It just shows OC OverCrowding=False Count=10, Entry=10, etc
Can I get this data?
Because I want to show them on the web, but I can’t figure out how to set it up.
Does deepstream have a plug-in or function switch setting to get this data for me to process further.
Actually, I was thinking about whether I could get this data in real time for display, or write it into a log so that I could read it in python, or some other way to get it
