/hack_pro_safety_doors_case

Primary LanguageJupyter Notebook

Safety Doors

Program for detecting people jammed by doors in trains.

Installation instruction

Install

Setup python environment

  • Install python3.7
sudo apt-get update

sudo apt-get install software-properties-common

sudo add-apt-repository ppa:deadsnakes/ppa
sudo apt-get update

sudo apt-get install python3-pip (not sure if this nessesary)
sudo apt-get install python3.6
sudo apt-get install python3.6-venv
  • Create virtual enviroment:
python3.6 -m venv venv
  • Activate environment
source venv/bin/activate
  • Install requirements:
venv/bin/pip install --upgrade pip && venv/bin/pip install -r requirements.txt

Usage

Analyze scene from .pcd file and generate output json

./scene_analyzer.py -i RES/example_clouds/10.pcd -o out.json

View scene with detected objects

./json_mapped.py -i RES/example_clouds/10.pcd -p out.json

VoiceAlert

Tested on Windows 10 64x

Install requirements:

  • SpeechRecognition
   pip install SpeachRecognition
  • PyAudio for Python 3.7 and Windows 64x
   pip install RES/PyAudio-0.2.11-cp37-cp37m-win_amd64.whl
  • Scikit-learn 0.24.2
   pip install -U scikit-learn

Usage

Set audio source. We can get all possible sound sources from the list speech_recognition.Microphone()

   import speech_recognition as sr
   ...
    with sr.Microphone(device_index=1) as source:
    ...

device_index specified the device from which the sound will be read