Skip to content

Latest commit

 

History

History
 
 

example_image_segmentation_tensorflow_lite


title: Image Segmentation (edgeTPU) ...

NNStreamer Edge-AI Apllication - Image Segmentation (edgeTPU)

Introduction

Image segmentation is the process of partitioning a digital image into multiple segments. The purpose of this app is show how to send the flatbuf to the edge device and perform inferences on the edgeTPU.

Graphical description of the pipeline

pipeline-img

How to run

  • Preparing nnstreamer for execution, refer this link
  • Install Prerequisites
    • Install flatbuffers runtime library
    $ sudo apt-get update
    $ sudo apt-get install libflatbuffers
    
    • Install GStreamer Data Protocol (libgstgdp.so)
      • For tizen device, download gst-plugins-bad*.rpm from here
      • For ubuntu, $ sudo apt-get install libgstreamer-plugins-bad1.0-0
  • Download model files
    $ cd $NNST_ROOT/bin/
    $ ./get-model.sh image-segmentation-tflite