Description
This plugin is performs object detetion using edgetpu on jetson nano.
Pre-requisites
- GStreamer-1.0 Development package
- GStreamer-1.0 Base Plugins Development package
- OpenCV Development package
- Jetpack 4.4
- Deepstream 5.0
Installation
-
Clone the Git Repository directory
-
Run
sudo make install -
Verify Installation
gst-inspect-1.0 dscoralinferExpect output
Factory Details:
Rank primary (256)
Long-name Coral Infer plugin for NVIDIA Deepstream
Klass Coral Infer plugin for NVIDIA Deepstream
Description Perform inference on Google coral usb accelerator
Author Stackfusion PVT LTD (https://www.parkzap.com)
Plugin Details:
Name nvdsgst_dscoralinfer
Description Inference plugin for google coral for NVIDIA Deepstream
Filename /usr/lib/aarch64-linux-gnu/gstreamer-1.0/deepstream/libnvdsgst_dscoralinfer.so
Version 5.0.0
License Proprietary
Source module dscoralinfer
Binary package Inference plugin for google coral for NVIDIA Deepstream
Origin URL https://parkzap.com/
GObject
+----GInitiallyUnowned
+----GstObject
+----GstElement
+----GstBaseTransform
+----GstCoralInfer
Pad Templates:
SRC template: 'src'
Availability: Always
Capabilities:
video/x-raw(memory:NVMM)
format: { (string)NV12, (string)RGBA, (string)I420 }
width: [ 1, 2147483647 ]
height: [ 1, 2147483647 ]
framerate: [ 0/1, 2147483647/1 ]
SINK template: 'sink'
Availability: Always
Capabilities:
video/x-raw(memory:NVMM)
format: { (string)NV12, (string)RGBA, (string)I420 }
width: [ 1, 2147483647 ]
height: [ 1, 2147483647 ]
framerate: [ 0/1, 2147483647/1 ]
Element has no clocking capabilities.
Element has no URI handling capabilities.
Pads:
SINK: 'sink'
Pad Template: 'sink'
SRC: 'src'
Pad Template: 'src'
Element Properties:
name : The name of the object
flags: readable, writable
String. Default: "coralinfer0"
parent : The parent of the object
flags: readable, writable
Object of type "GstObject"
qos : Handle Quality-of-Service events
flags: readable, writable
Boolean. Default: false
unique-id : Unique ID for the element. Can be used to identify output of the element
flags: readable, writable
Unsigned Integer. Range: 0 - 4294967295 Default: 15
processing-width : Width of the input buffer to algorithm
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 300
processing-height : Height of the input buffer to algorithm
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 300
full-frame : Enable to process full frame or disable to process objects detectedby primary detector
flags: readable, writable
Boolean. Default: true
gpu-id : Set GPU Device ID
flags: readable, writable, changeable only in NULL or READY state
Unsigned Integer. Range: 0 - 4294967295 Default: 0
model-path : path of model used by edge-tpu
flags: readable, writable
String. Default: null
label-path : path of label used by edge-tpu
flags: readable, writable
String. Default: null
threshold : threshold of object to be detected by edge-tpu
flags: readable, writable
Float. Range: 0 - 1 Default: 0.7
config-file-path : Absolute pathname of configuration file for the coralinfer element
flags: readable, writable
String. Default: null
detected-max-h : Default max height of detected class
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0
detected-max-w : Default max width of detected class
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0
detected-min-h : Default min height of detected class
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0
detected-min-w : Default min width of detected class
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0
enable : enable class by defaultby primary detector
flags: readable, writable
Boolean. Default: false
eps : eps values for overlapping, to be used by cv::groupRectangles edge-tpu
flags: readable, writable
Double. Range: 0 - 1.797693e+308 Default: 0
group-threshold : Min number of rectangle to consider a cluster to be used by cv::groupRectangles
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0Usage
gst-launch-1.0 uridecodebin uri=file://<video_path> ! m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=720 ! dscoralinfer config-file-path="<config_file_path>" ! nvvideoconvert ! nvtracker tracker-width=640 tracker-height=320 ll-lib-file=/opt/nvidia/deepstream/deepstream-5.0/lib/libnvds_mot_klt.so ll-config-file=<tracker_config_path> ! nvdsosd ! nvvideoconvert ! fpsdisplaysink name=fpssink text-overlay=true video-sink=xvimagesink sync=0Example:
gst-launch-1.0 uridecodebin uri=file:///home/stackfusion/pipline/sample_short.mp4 ! m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=720 ! dscoralinfer config-file-path=resource/sample_config.txt ! nvvideoconvert ! nvtracker tracker-width=640 tracker-height=320 ll-lib-file=/opt/nvidia/deepstream/deepstream-5.0/lib/libnvds_mot_klt.so ll-config-file=/resource/tracker_config_1.yml ! nvdsosd ! nvvideoconvert ! fpsdisplaysink name=fpssink text-overlay=true video-sink=xvimagesink sync=0Running a Benchmark
gst-launch-1.0 uridecodebin uri=file:///home/stackfusion/pipline/sample_short.mp4 ! m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=720 ! dscoralinfer config-file-path="resource/sample_config.txt" ! nvvideoconvert ! nvtracker tracker-width=640 tracker-height=320 ll-lib-file=/opt/nvidia/deepstream/deepstream-5.0/lib/libnvds_mot_klt.so ll-config-file=resource/tracker_config_1.yml ! nvdsosd ! nvvideoconvert ! fpsdisplaysink name=fpssink text-overlay=true video-sink=xvimagesink sync=0Results
Device : Jetson Nano (Developer kit version)
Jetpack : 4.4DP
Deepstream : 5.0
Model : SSD Movilenet v2
| Frame rate | 54 fps |
| Ram Usage | 71Mb |
| Cpu Usage | 30% |
Properties:
name : The name of the object
flags: readable, writable
String. Default: "coralinfer0"
parent : The parent of the object
flags: readable, writable
Object of type "GstObject"
qos : Handle Quality-of-Service events
flags: readable, writable
Boolean. Default: false
unique-id : Unique ID for the element. Can be used to identify output of the element
flags: readable, writable
Unsigned Integer. Range: 0 - 4294967295 Default: 15
processing-width : Width of the input buffer to algorithm
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 300
processing-height : Height of the input buffer to algorithm
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 300
full-frame : Enable to process full frame or disable to process objects detectedby primary detector
flags: readable, writable
Boolean. Default: true
gpu-id : Set GPU Device ID
flags: readable, writable, changeable only in NULL or READY state
Unsigned Integer. Range: 0 - 4294967295 Default: 0
model-path : path of model used by edge-tpu
flags: readable, writable
String. Default: null
label-path : path of label used by edge-tpu
flags: readable, writable
String. Default: null
threshold : threshold of object to be detected by edge-tpu
flags: readable, writable
Float. Range: 0 - 1 Default: 0.7
config-file-path : Absolute pathname of configuration file for the coralinfer element
flags: readable, writable
String. Default: null
detected-max-h : Default max height of detected class
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0
detected-max-w : Default max width of detected class
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0
detected-min-h : Default min height of detected class
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0
detected-min-w : Default min width of detected class
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0
enable : enable class by defaultby primary detector
flags: readable, writable
Boolean. Default: false
eps : eps values for overlapping, to be used by cv::groupRectangles edge-tpu
flags: readable, writable
Double. Range: 0 - 1.797693e+308 Default: 0
group-threshold : Min number of rectangle to consider a cluster to be used by cv::groupRectangles
flags: readable, writable
Integer. Range: 1 - 2147483647 Default: 0Sample
Vehicle counting sample pipeline
With Coral USB Accelerator
gst-launch-1.0 uridecodebin uri=file:///home/stackfusion/pipline/sample_short.mp4 ! m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=720 ! dscoralinfer config-file-path="resource/sample_config.txt" ! nvvideoconvert ! nvtracker tracker-width=640 tracker-height=320 ll-lib-file=/opt/nvidia/deepstream/deepstream-5.0/lib/libnvds_mot_klt.so ll-config-file=resource/tracker_config_1.yml ! nvdsosd ! nvvideoconvert ! fpsdisplaysink name=fpssink text-overlay=true video-sink=xvimagesink sync=0Without Coral USB Accelerator
gst-launch-1.0 filesrc location=/home/stackfusion/pipline/ds-edgetpu-infer/sample_short.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=720 ! nvinfer config-file-path= /opt/nvidia/deepstream/deepstream-5.0/samples/configs/deepstream-app/config_infer_primary_nano.txt batch-size=1 unique-id=1 ! nvvideoconvert ! nvtracker tracker-width=640 tracker-height=320 ll-lib-file=/opt/nvidia/deepstream/deepstream-5.0/lib/libnvds_mot_klt.so ll-config-file=resource/tracker_config_1.yml ! nvdsanalytics config-file=resource/config_nvdsanalytics.txt ! nvdsosd ! nvvideoconvert ! fpsdisplaysink name=fpssink text-overlay=true video-sink=xvimagesink sync=0Memory Usage with Coral Plugin
Vehicle Counting and Classification using Edge AI device in realtime with Coral TPU + Jetson Nano
Memory Usage without Coral Plugin
Vehicle Counting and Classification using Edge AI device in realtime
