Merge pull request #139 from tommywu052/master

Factory AI vison on WSL2 GPU enabled.

factory-ai-vision/rtsp-generator/README.md

@@ -0,0 +1,35 @@
+<h1 align="center">RTSP Endpoint with Gstreamer 👋</h1>
+<p>
+ How to generate RTSP endpoint with GStreamer.
+</p>
+
+
+
+## GStreamer Installation
+
+```sh
+sudo apt-get install libgstreamer1.0-0 gstreamer1.0-plugins-base gstreamer1.0-plugins-good gstreamer1.0-plugins-bad gstreamer1.0-plugins-ugly gstreamer1.0-libav gstreamer1.0-doc gstreamer1.0-tools
+sudo apt-get install gstreamer1.0-plugins-base-apps
+sudo apt-get install gir1.2-gst-rtsp-server-1.0
+sudo apt-get install python-gst-1.0 python3-gst-1.0
+sudo apt-get install python3-opencv
+
+```
+
+## Usage
+
+```sh
+Download gst-rtsp.py to current folder
+
+Modify Line 14 for your own video or url 
+> - self.cap = cv2.VideoCapture("your own video file or url ")
+python3 gst-rtsp.py
+```
+
+## Author
+
+👤 **Tommy Wu**
+
+* Github: [@tommywu052](https://github.com/tommywu052)
+
+## Thanks you

factory-ai-vision/rtsp-generator/gst-rtsp.py

@@ -0,0 +1,72 @@
+#!/usr/bin/env python3
+
+import cv2
+import gi
+
+gi.require_version('Gst', '1.0')
+gi.require_version('GstRtspServer', '1.0')
+from gi.repository import Gst, GstRtspServer, GObject
+
+import netifaces as ni
+ni.ifaddresses('eth0')
+ip = ni.ifaddresses('eth0')[ni.AF_INET][0]['addr']
+
+class SensorFactory(GstRtspServer.RTSPMediaFactory):
+    def __init__(self, **properties):
+        super(SensorFactory, self).__init__(**properties)
+        self.cap = cv2.VideoCapture("taiwan_culture.mp4")
+        self.number_frames = 0
+        self.fps = 30
+        self.duration = 1 / self.fps * Gst.SECOND  # duration of a frame in nanoseconds
+        self.launch_string = 'appsrc name=source is-live=true block=true format=GST_FORMAT_TIME ' \
+                             'caps=video/x-raw,format=BGR,width=1280,height=720,framerate={}/1 ' \
+                             '! videoconvert ! video/x-raw,format=I420 ' \
+                             '! x264enc speed-preset=ultrafast tune=zerolatency ' \
+                             '! rtph264pay config-interval=1 name=pay0 pt=96'.format(self.fps)
+
+    def on_need_data(self, src, lenght):
+        if self.cap.isOpened():
+            ret, frame = self.cap.read()
+            if ret:
+                data = frame.tostring()
+                #print (data)
+                buf = Gst.Buffer.new_allocate(None, len(data), None)
+                buf.fill(0, data)
+                buf.duration = self.duration
+                timestamp = self.number_frames * self.duration
+                buf.pts = buf.dts = int(timestamp)
+                buf.offset = timestamp
+                self.number_frames += 1
+                retval = src.emit('push-buffer', buf)
+                print('pushed buffer, frame {}, duration {} ns, durations {} s'.format(self.number_frames,
+                                                                                       self.duration,
+                                                                                       self.duration / Gst.SECOND))
+                if retval != Gst.FlowReturn.OK:
+                    print(retval)
+
+    def do_create_element(self, url):
+        return Gst.parse_launch(self.launch_string)
+
+    def do_configure(self, rtsp_media):
+        self.number_frames = 0
+        appsrc = rtsp_media.get_element().get_child_by_name('source')
+        appsrc.connect('need-data', self.on_need_data)
+
+
+class GstServer(GstRtspServer.RTSPServer):
+    def __init__(self, **properties):
+        super(GstServer, self).__init__(**properties)
+        self.factory = SensorFactory()
+        self.factory.set_shared(True)
+        self.get_mount_points().add_factory("/test", self.factory)
+        self.attach(None)
+
+
+GObject.threads_init()
+Gst.init(None)
+
+server = GstServer()
+print("Running as - rtsp://"+ip+":8554/test")
+
+loop = GObject.MainLoop()
+loop.run()

factory-ai-vision/wsl2/README.md

@@ -0,0 +1,78 @@
+# RUNNING FACTORY AI ON WSL2 with GPU Acceleration Enabled
+
+
+1.  WSL2 Environment Setup :
+
+    A.  Surface Book2 with GTX 1060
+
+    B.  Windows Insider Preview Build 20150 or higher :  [register for
+        the Windows Insider
+        Program](https://insider.windows.com/getting-started/#register).
+
+    C.  Download Nvidia Driver :
+        <https://developer.nvidia.com/45541-gameready-win10-dch-64bit-international>
+
+    D.   [Enable WSL
+        2](https://docs.microsoft.com/en-us/windows/wsl/install-win10) and [install
+        a glibc-based
+        distribution](https://docs.microsoft.com/en-us/windows/wsl/install-win10#install-your-linux-distribution-of-choice) (like
+        Ubuntu or Debian). I choose Ubuntu 18.04
+
+    E.  Check this
+        [guide](https://docs.nvidia.com/cuda/wsl-user-guide/index.html#setting-containers)
+        to install Nvidia container. (Should install this before IoT
+        Edge installation) .
+
+2.  Install IoT Edge and Configure your device on IoT Hub by the
+    [doc](https://docs.microsoft.com/en-us/azure/iot-edge/how-to-install-iot-edge-linux).
+
+![](media/image3.png)
+
+3.  Steps to run AI workload for Web Module & Inference Module --
+
+    A.  Start the previous distro environment
+
+> ![](media/image4.png)
+    B.  Clone the
+    [start.sh](https://github.com/tommywu052/azure-intelligent-edge-patterns/blob/master/factory-ai-vision/wsl2/start.sh)
+    on the root folder.
+
+    C.  cd \~ & ./start.sh
+
+> ![](media/image5.png)
+
+    D.  Check your module status :
+
+> ![](media/image6.png)
+
+    E.  nano the /etc/docker/daemon.json as
+
+> ![](media/image7.png)
+
+    F.  systemctl restart docker & iotedge restart
+
+```{=html}
+```
+4.  Testing Process :
+
+    A.  Check WebModule is running :
+
+         iotedge logs -f WebModule \--tail 500
+
+    B.  Go to <http://localhost:8080/> and add a new location.
+
+> ![](media/image9.png)
+
+    C.  Choose "demo Pretrained Detection " and Configure
+
+    D.  Check the Right-hand side Configuration about GPU(accelerated) :
+        around 10-15 ms , compared with CPU around 300-500 ms.
+
+![](media/image11.png)
+
+    E.  Verify the logs on inference module : some onnxruntime warning but
+        initializer scalepreprocessor scale.
+
+   > ![](media/image12.png)
+
+    F.  Done and Enjoy your WSL2 !

factory-ai-vision/wsl2/daemon.json

@@ -0,0 +1,10 @@
+{
+  "default-runtime": "nvidia",
+  "runtimes": {
+        "nvidia": {
+            "path": "nvidia-container-runtime",
+            "runtimeArgs": []
+        }
+    },
+  "dns": ["8.8.8.8", "8.8.4.4"]
+}

factory-ai-vision/wsl2/media/README.md

@@ -0,0 +1 @@
+

factory-ai-vision/wsl2/start.sh

@@ -0,0 +1,2 @@
+sudo daemonize /usr/bin/unshare --fork --pid --mount-proc /lib/systemd/systemd --system-unit=basic.target
+exec sudo nsenter -t $(pidof systemd) -a su - $LOGNAME