README

*****************************************************************************
* Copyright (c) 2019-2020 NVIDIA Corporation.  All rights reserved.
*
* NVIDIA Corporation and its licensors retain all intellectual property
* and proprietary rights in and to this software, related documentation
* and any modifications thereto.  Any use, reproduction, disclosure or
* distribution of this software and related documentation without an express
* license agreement from NVIDIA Corporation is strictly prohibited.
*****************************************************************************

Prequisites:

Please follow instructions in the apps/sample_apps/deepstream-app/README on how
to install the prequisites for Deepstream SDK, the DeepStream SDK itself and the
apps.

You must have the following development packages installed
   GStreamer-1.0
   GStreamer-1.0 Base Plugins
   GStreamer-1.0 gstrtspserver
   X11 client-side library

To install these packages, execute the following command:
   sudo apt-get install libgstreamer-plugins-base1.0-dev libgstreamer1.0-dev \
   libgstrtspserver-1.0-dev libx11-dev

Pipeline:
H264/JPEG-->decoder-->tee -->| -- (batch size) ------>|-->streammux--> nvinfer--> nvsegvisual ----> nvosd --> |---> encode --->filesink

Compilation Steps:
  $ cd apps/deepstream-segmentation-test/
  $ make

the usr_input.txt
=================
this is the user input file to define the input image batch size, output display height and width, stream directories. Please follow exact format as below example: i.e. no less or space, no variable name change. for example, stream1 can not be images1, etc. If user does not want change the value, then following setting will be the default. the program will be run forever unless user will shut if off.  

example 
===========
batch_size=8
width=1280
height=720
stream1=/opt/nvidia/deepstream/deepstream-5.0/sources/apps/sample_apps/deepstream-segmentation-analytics/images1
stream2=/opt/nvidia/deepstream/deepstream-5.0/sources/apps/sample_apps/deepstream-segmentation-analytics/images2
pro_per_sec=40
no_streams=3
production=1

user defined parameters
=======================

batch_size  : how many total images each time to load for a stram directory
width       : output image width
height      : output image height
stream1     : path to image1 directory
stream2     : path to image2 directory
pro_per_sec : how many seconds to wait before a new segmentation run
no_streams  : number of stream directories
production  : 1 for real production env. 0 for Nvidia helm-chart env.
 

Example command to run:
=======================

For binary segmentation:

 $ ./deepstream-segmentation-analytics -c dstest_segmentation_config_industrial.txt -i usr_input.txt

For multi-class semantic segmentation: 

 $ ./deepstream-segmentation-analytics -c dstest_segmentation_config_semantic.txt -i usr_input.txt


Other Note
==========
for Helm-Chart env., segmentation run generate out.jpg for the mask ground truth 
for the producton env., the mask directory will have the all the mask ground truth pictures and input directory will save the input images in case to be used for the retrain purpose