Patent · US Active

Traffic-light detection and classification using computer vision and deep learning

US10185881B2 · kind B2 · utility

3Cited by
0References
20Claims
0Family size

Assignee

Inventors

Key dates

Filing dateNov 23, 2016
Grant dateJan 22, 2019
Priority date
Expiry dateJan 4, 2037

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06T2207/20024
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

A method is disclosed for detecting and classifying one or more traffic lights. The method may include converting an RGB frame to an HSV frame. The HSV frame may be filtered by at least one threshold value to obtain at least one saturation frame. At least one contour may be extracted from the at least one saturation frame. Accordingly, a first portion of the RGB may be cropped in order to encompass an area including the at least one contour. The first portion may then be classified by an artificial neural network to determined whether the first portion corresponds to a not-a-traffic-light class, a red-traffic-light class, a green-traffic-light class, a yellow-traffic-light class, or the like.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.