Skip to content

Detect and Classify Red Traffic Signs (Intredit/Prohibition Traffic Signs)

Notifications You must be signed in to change notification settings

Chubek/intresignia

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Intresignia

CircleCI

Status

  • Verson 0.0.1beta [Deprecated]-> Detection done.
  • Verson 0.0.2beta [Deprecated]-> Improved magnificantly, with a lot of stuff added, including preliminary classification and loads of parameters.
  • Verson 0.0.3beta [Current]-> Recognition and Classification.

Quick Test

Test 0.0.3beta here:

Open In Colab

What is this?

This is a Python package that uses color thresholding and other classical vision methods through OpenCV to detect a subset of intredit (prohibition) traffic signs. It will detect and classify any sign that is red.

How to Use?

  1. This package is not hosted on PyPi so don't try installing it with pip like that. Just do this:
python3.10 virtualenv venv
source venv/bin/activate (on Linux) or venv/vin/activate.ps1 (Windows)
python3.10 pip install git+https://github.com/chubek/intresignia.git

This will install the package. Mind you that you need to install Python 3.10. It does not use any of its feaures yet but it will pretty soon.

  1. Create a new Python file and:
from intresignia import Settings, intresignia_detect
import cv2
from pprint import pprint

st = Settings()

det, ssim_scores, coords, color_isolated = intresignia_detect("/path/to/img.png", st)

pprint(ssim_scores)
pprint(coords)


cv2.imshow('Detected Signs', det)
cv2.waitKey(0)
cv2.destroyAllWindows()

An exhaustive list of settings + grid search to find the best params using grid_search.py will be added to this readme pretty soon. Meanwhile please use the help(Settings).

Any questions regarding this library should be directed at Chubak#7400.