Clearbot Vision Module

Setup guide

Requirements

• Python 3.6
• USB Camera/webcam

Setup environment

Using virtualevn:

If you don't have virtualenv installed:

pip install virtualenv --user

In the project directory, setup the virtual environment:

virtualenv .venv -p python3

Use the environment while working on the project:

Linux:

source .venv/bin/activate

Windows (using Command Prompt)

cd .venv/Scripts
activate.bat
cd ..\..

Install Python dependencies using:

pip install -r requirements.txt

Download model files

There are two models in this project: YOLO and Mask R-CNN.

The model folders must be created in ml/yolo/model and ml/detectron/model.

Please find the models in the release v0.2.

Run the module

python main.py

Please refer to the options available for the command below.

Note: If the above command does not run, make sure the virtual environment is activated.

Command Line Reference

Option	Value	Description
-v	boolean (True/False)	Show video frame output. Default: false (only available for yolo model)
--debug	boolean (True/False	Switch to debug mode for logs: Default: false (only available for the yolo model)
-m	detectron/yolo	Model type to use for detection. Default: full
-s	full/tiny	Model type to use for detection. Default: full (only available for the YOLO model)