TY - GEN
T1 - The Improved Security System in Smart Wheelchairs for Detecting Stair Descent using Image Analysis
AU - Utaminingrum, Fitri
AU - Johan, Ahmad Wali Satria Bahari
AU - Sari, Yuita Arum
AU - Somawirata, I. Komang
AU - A Olaode, Abass Abolarinwa
N1 - Publisher Copyright:
© 2021 ACM.
PY - 2021/2/23
Y1 - 2021/2/23
N2 - A smart wheelchair requires a security system for its users to feel safe and comfortable. The process of observing road conditions is one of the solutions to maintaining user safety, which one of these hurdles can be a sudden transition of the situation in surface road height level for example, such as a descending staircase. Integration system for safety in smart wheelchairs consists of three main parts, namely input (Camera), output (Driver motor Left and Right), and main processing (Mini PC). The proposed research will be carried out stair decent detection using a Gray Level Co-occurrence matrix (GLCM) algorithm method as an extraction feature algorithm. The usage of GLCM methods can be applied to images that have textures. While if we look at the descent of the stairs also has a different texture when compared to the usual floor. Support Vector Machine (SVM) is used for the classification of stairs descent and floor. SVM algorithms have advantageous in it is effortless and strong consistency of implementation in classification. In this research propose combination methods between the texture features using GLCM and classification method using SVM to obtain effective detection stairs descent and floor.The proposed method by setting the GLCM parameter with a value of d = 1 and θ = 135o, and SVM classification using the Radial Basis Function Kernel (RBF Kernel) has an accuracy of 87 for detecting the stair descent with relatively fast computation time equal to 0.007 second.
AB - A smart wheelchair requires a security system for its users to feel safe and comfortable. The process of observing road conditions is one of the solutions to maintaining user safety, which one of these hurdles can be a sudden transition of the situation in surface road height level for example, such as a descending staircase. Integration system for safety in smart wheelchairs consists of three main parts, namely input (Camera), output (Driver motor Left and Right), and main processing (Mini PC). The proposed research will be carried out stair decent detection using a Gray Level Co-occurrence matrix (GLCM) algorithm method as an extraction feature algorithm. The usage of GLCM methods can be applied to images that have textures. While if we look at the descent of the stairs also has a different texture when compared to the usual floor. Support Vector Machine (SVM) is used for the classification of stairs descent and floor. SVM algorithms have advantageous in it is effortless and strong consistency of implementation in classification. In this research propose combination methods between the texture features using GLCM and classification method using SVM to obtain effective detection stairs descent and floor.The proposed method by setting the GLCM parameter with a value of d = 1 and θ = 135o, and SVM classification using the Radial Basis Function Kernel (RBF Kernel) has an accuracy of 87 for detecting the stair descent with relatively fast computation time equal to 0.007 second.
KW - and Smart wheelchair
KW - GLCM
KW - RBF Kernel
KW - Stair descent
KW - SVM
UR - https://www.scopus.com/pages/publications/85112294437
U2 - 10.1145/3457784.3457808
DO - 10.1145/3457784.3457808
M3 - Conference contribution
AN - SCOPUS:85112294437
T3 - ACM International Conference Proceeding Series
SP - 150
EP - 154
BT - 2021 10th International Conference on Software and Computer Applications, ICSCA 2021
PB - Association for Computing Machinery
T2 - 10th International Conference on Software and Computer Applications, ICSCA 2021
Y2 - 23 February 2021 through 26 February 2021
ER -