Guide: Testing Pallet Defect Detection (PDD) Application Using Balluff SDK#

This guide explains how to create a custom Docker image based on the Intel DL Streamer Pipeline Server, with Balluff SDK and Gencamsrc support. It supports Balluff, Basler, and other GenICam-compatible cameras over USB and GigE interfaces.

Note: You may observe a watermark in the camera feed when testing with a non-Balluff camera, as it is the free version.

Prerequisites#

System Requirements

Cloning and building the docker image#

Step 1: Base Image and User Setup#

Download the edge-ai-libraries source and go to dlstreamer-pipeline-server folder

git clone https://github.com/open-edge-platform/edge-ai-libraries.git
cd edge-ai-libraries/microservices/dlstreamer-pipeline-server

Step 2: Create the Docker Image#

Create a Docker file named BalluffDockerfile inside your dlstreamer-pipeline-server directory with the following content.

FROM intel/dlstreamer-pipeline-server:3.1.0-ubuntu24

USER root

RUN apt-get update && apt-get install -y wget gnupg gstreamer1.0-plugins-base libxcb-cursor0 make autoconf vim libgstreamer-plugins-base1.0-dev libgstreamer1.0-dev g++-11 g++ && apt-get clean && rm -rf /var/lib/apt/lists/*

RUN update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-11 100

COPY ./thirdparty/install_gencamsrc_gstreamer_plugin.sh /home/pipeline-server/install_gencamsrc_gstreamer_plugin.sh

COPY ./plugins/camera/src-gst-gencamsrc /home/pipeline-server/src-gst-gencamsrc

RUN chmod +x /home/pipeline-server/src-gst-gencamsrc/setup.sh
RUN chmod +x /home/pipeline-server/src-gst-gencamsrc/autogen.sh
RUN chmod +x /home/pipeline-server/install_gencamsrc_gstreamer_plugin.sh
RUN /home/pipeline-server/install_gencamsrc_gstreamer_plugin.sh

# For Ubuntu24 intel/dlstreamer-pipeline-server:3.1.0-ubuntu24 base image
RUN apt-get update && apt-get install -y libwxgtk-webview3.2-dev

# For Ubuntu 22 with intel/dlstreamer-pipeline-server:3.1.0-ubuntu22, uncomment the line below and comment the above line
# RUN apt-get update && apt-get install -y libwxgtk-webview3.0-gtk3-dev

RUN mkdir /home/pipeline-server/Balluff_Impact_Acquire_V3
RUN wget https://assets-2.balluff.com/mvIMPACT_Acquire/3.0.0/install_mvGenTL_Acquire.sh -P /home/pipeline-server/Balluff_Impact_Acquire_V3
RUN wget https://assets-2.balluff.com/mvIMPACT_Acquire/3.0.0/mvGenTL_Acquire-x86_64_ABI2-3.0.0.tgz -P /home/pipeline-server/Balluff_Impact_Acquire_V3
RUN cd Balluff_Impact_Acquire_V3 && chmod +x ./install_mvGenTL_Acquire.sh && ./install_mvGenTL_Acquire.sh -u3v -gev -u

ENV MVIMPACT_ACQUIRE_DIR="/opt/mvIMPACT_Acquire" \
    MVIMPACT_ACQUIRE_DATA_DIR="/opt/mvIMPACT_Acquire/data" \
    GENICAM_ROOT="/opt/mvIMPACT_Acquire/runtime" \
    MVIMPACT_ACQUIRE_FAVOUR_SYSTEMS_LIBUSB="1" \
    GENICAM_GENTL64_PATH=/opt/Impact_Acquire/lib/x86_64 \
    GST_PLUGIN_PATH=/opt/intel/dlstreamer/lib:/opt/intel/dlstreamer/gstreamer/lib/gstreamer-1.0:/opt/intel/dlstreamer/gstreamer/lib/:/usr/lib/x86_64-linux-gnu/gstreamer-1.0:/usr/local/lib/gstreamer-1.0

USER intelmicroserviceuser

Step 3: Build the Docker Image#

Run the following command to build the image:

docker build -t intel/dlstreamer-pipeline-server:3.1.0-ubuntu24-gencamsrc-balluff -f BalluffDockerfile .

This command builds your Docker image using the steps defined above.

Step 4: Verify the Image#

After the build completes, update .env and start the container:

update .env DLSTREAMER_PIPELINE_SERVER_IMAGE=intel/dlstreamer-pipeline-server:3.1.0-ubuntu24-gencamsrc-balluff

docker compose up -d

Step 5: Run a test pipeline and dump the camera output into a file in the /tmp directory#

Note down serial number of the balluff camera and update <balluff-camera-serial> in the following command

docker exec -it dlstreamer-pipeline-server bash
$ gst-launch-1.0 gencamsrc serial=<balluff-camera-serial> pixel-format=bayerrggb name=source ! bayer2rgb ! videoscale ! video/x-raw, width=1920,height=1080 ! videoconvert ! queue ! jpegenc ! avimux ! filesink location=/tmp/gencam_balluff_output.avi

Verify that the /tmp/gencam_balluff_output.avi has the captured content

Deploying the Pallet Defect Detection (PDD) Application Using live camera#

This guide provides detailed, step-by-step instructions for setting up and deploying the Pallet Defect Detection (PDD) pipeline to use the Balluff or Basler camera connected over USB or GigE.
It covers environment setup, configuration updates, and validation steps to ensure a successful deployment.

Step 1: Set Up the Environment#

git clone https://github.com/open-edge-platform/edge-ai-suites.git
cd edge-ai-suites/manufacturing-ai-suite/industrial-edge-insights-vision
cp .env_pallet_defect_detection .env

Step 2: Configure the .env File#

Update the .env file with the newly created image as below and modify any other required variables.

DLSTREAMER_PIPELINE_SERVER_IMAGE=intel/dlstreamer-pipeline-server:3.1.0-ubuntu24-gencamsrc-balluff

Step 3: Run the Setup Script#

Execute the setup script to initialize project directories and configurations.

./setup.sh

Step 4: Update the Pipeline Configuration#

Update the pipeline in ./apps/pallet-defect-detection/configs/pipeline-server-config.json to use the camera:

{
    "name": "pallet_defect_detection",
    "source": "gstreamer",
    "queue_maxsize": 50,
    "pipeline": "gencamsrc serial=<camera id> pixel-format=bayerrggb name=source ! bayer2rgb ! videoscale ! video/x-raw, width=640, height=480 ! videoconvert ! gvadetect name=detection model-instance-id=inst0 ! gvametaconvert add-empty-results=true name=metaconvert ! queue ! gvafpscounter ! gvawatermark ! appsink name=destination"
}

Replace with balluff/balser camera id connected over the USB or GigE

Step 5: Configure docker-compose.yml (Optional, if testing with a GigE network camera)#

When testing with a GigE camera, you need to adjust the docker-compose.yml configuration for all the services. Follow these steps:

Add network_mode set to "host".
Remove the networks section.

The configuration for each service should look like this:

services:
  service_name:
    .
    .
    network_mode: "host"
    # networks:
    #   - mraas

Additionally, add the following entries to the /etc/hosts file on the host machine:

0.0.1       dlstreamer-pipeline-server
0.0.1       prometheus
0.0.1       mediamtx-server
0.0.1       mraas-minio
0.0.1       otel-collector
0.0.1       mqtt-broker
0.0.1       model_registry

Step 6: Launch the Containers#

Start all required services using Docker Compose:

docker compose up -d

Step 7: Modify the Payload File#

Edit edge-ai-suites/manufacturing-ai-suite/industrial-edge-insights-vision/apps/pallet-defect-detection/payload.json and remove the source section so that it looks like this:

[
    {
        "pipeline": "pallet_defect_detection",
        "payload": {
            "destination": {
                "frame": {
                    "type": "webrtc",
                    "peer-id": "pdd",
                    "overlay": false
                }
            },
            "parameters": {
                "detection-properties": {
                    "model": "/home/pipeline-server/resources/models/pallet-defect-detection/deployment/Detection/model/model.xml",
                    "device": "CPU"
                }
            }
        }
    }
]

Step 8: Start the Sample Pipeline#

Run the sample script to start the pipeline:

./sample_start.sh -p pallet_defect_detection

Step 9: Access the Web Interface#

Open a browser and navigate to:

https://<HOST_IP>/mediamtx/pdd/

Replace <HOST_IP> with the IP address configured in your .env file.

Troubleshooting#

For initial configuration and advanced configuration of the Balluff camera, please use the company-provided visualization tool ImpactAcquire, which is part of the Balluff SDK.
Instructions to install the Balluff SDK on the host can be found in the Balluff SDK Installation Guide.