Sistem Deteksi Lubang pada Pedesterian dengan Teknik Pengolahan Citra

Intan Sari Areni; Indrabayu Amirullah; Ingrid Nurtanio; Anugrayani Bustamin; Ahmad Rifaldi

doi:10.25042/jpe.112019.04

Intan Sari Areni Departemen Teknik Elektro, Fakultas Teknik, Universitas Hasanuddin
Indrabayu Amirullah Departemen Teknik Informatika, Fakultas Teknik, Universitas Hasanuddin
Ingrid Nurtanio Departemen Teknik Informatika, Fakultas Teknik, Universitas Hasanuddin
Anugrayani Bustamin Departemen Teknik Informatika, Fakultas Teknik, Universitas Hasanuddin
Ahmad Rifaldi Departemen Teknik Informatika, Fakultas Teknik, Universitas Hasanuddin

DOI: https://doi.org/10.25042/jpe.112019.04

Keywords: Tuna netra, blob analysis, deteksi lubang, HSV

Abstract

Pothole Detection System on Pedesterian using Image Processing Techniques. The pedestrian areas in Indonesia are still far from optimal in facilitating the users or the pedestrians. Potholed pedestrian areas are found in many parts of the street. This issue can harm pedestrians, especially blind people. For this reason, research has been carried out to create a system that can detect and estimate hole distances by processing images using mono cameras that can help blind people. The methods used to detect holes are the Threshold + Blob Analysis method and the HSV method. The obtained results indicate the level of accuracy of hole detection using the Threshold + Blob Analysis method is better than the HSV method. The average accuracy level of Threshold + Blob Analysis is 88.91%, while for the HSV method is 86.82%.