TY - GEN
T1 - An Obstacle Detection Method for Visually Impaired People Based on Semantic Segmentation
AU - Chen, Zhuo
AU - Liu, Xiaoming
AU - Liu, Dan
AU - Tang, Xiaoqing
AU - Huang, Qiang
AU - Arai, Tatsuo
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.
PY - 2024
Y1 - 2024
N2 - Using low-cost visual sensors to assist indoor and outdoor navigation is an important method to solve the problem of visually impaired people living and going out. To this end, we proposed an obstacle-detection method for visually impaired people based on semantic segmentation. We use the semantic segmentation method to determine which targets in the camera view field need to be noticed and use the related information to establish a real-time local map. At the same time, we propose a method to fuse semantic information with local point clouds, achieving obstacle detection based on probability fusion. Finally, the distance between the interested target and the camera will be returned and sent to the user. The proposed method can achieve visual navigation with more than ten frames per second (fps), lower than 0.3 m detection accuracy, and smaller than 4 MB generated model, which is also compatible with multiple cameras and control terminals.
AB - Using low-cost visual sensors to assist indoor and outdoor navigation is an important method to solve the problem of visually impaired people living and going out. To this end, we proposed an obstacle-detection method for visually impaired people based on semantic segmentation. We use the semantic segmentation method to determine which targets in the camera view field need to be noticed and use the related information to establish a real-time local map. At the same time, we propose a method to fuse semantic information with local point clouds, achieving obstacle detection based on probability fusion. Finally, the distance between the interested target and the camera will be returned and sent to the user. The proposed method can achieve visual navigation with more than ten frames per second (fps), lower than 0.3 m detection accuracy, and smaller than 4 MB generated model, which is also compatible with multiple cameras and control terminals.
KW - Assistance for Visually Impaired People
KW - Obstacle Detection
KW - Semantic Segmentation
UR - https://www.scopus.com/pages/publications/85187710202
U2 - 10.1007/978-981-97-0885-7_3
DO - 10.1007/978-981-97-0885-7_3
M3 - Conference contribution
AN - SCOPUS:85187710202
SN - 9789819708840
T3 - Communications in Computer and Information Science
SP - 28
EP - 33
BT - Cognitive Computation and Systems - 2nd International Conference, ICCCS 2023, Revised Selected Papers
A2 - Sun, Fuchun
A2 - Li, Jianmin
PB - Springer Science and Business Media Deutschland GmbH
T2 - 2nd International Conference on Cognitive Computation and Systems, ICCCS 2023
Y2 - 14 October 2023 through 15 October 2023
ER -