Staircase detection for safe mobility of visually impaired people

Vision is extremely important to human beings, due to the fact that it perceives and interprets everything around by simply looking at it and its visual features. But some individuals have some kind of visual impairment and face many difficulties in their day-to-day life. They need assistance for na...

Full description

Saved in:
Bibliographic Details
Main Authors: Madake, Jyoti, Bhatlawande, Shripad, Shilaskar, Swati, Deshpande, Piyush, Deshpande, Viraj
Format: Conference Proceeding
Language:English
Subjects:
Algorithms
Classifiers
Decision trees
Feature extraction
Machine learning
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Vision is extremely important to human beings, due to the fact that it perceives and interprets everything around by simply looking at it and its visual features. But some individuals have some kind of visual impairment and face many difficulties in their day-to-day life. They need assistance for navigation. This paper proposes a cane-based system to detect the presence of a staircase and alert the user. For the detection of stairs, a machine learning-based mode is developed. In this work SIFT is used for feature extraction. PCA is used for dimensionality reduction. 5 classifier algorithms are compared to classify the staircase. The classification accuracy for the classifiers is KNN 79.89%, Random Forest 95.54%, Logistic Regression 61.62%, SVM 72.29%, Decision Tree 82.99%.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0161151