DEEP STAIRS DETECTION BASED ON CONVOLUTIONAL NEURAL NETWORKS FOR VISUALLY IMPAIRED PERSONS
DOI:
https://doi.org/10.11113/aej.v15.23383Keywords:
Visual impairment, Assistive technology, Convolutional neural networks, Stair fall risk, Deep learningAbstract
In 2022, an alarming 2.2 billion people worldwide suffered from some form of vision impairment, with 237 million individuals experiencing moderate to severe visual impairment, putting them at a heightened risk of accidents and injuries. Clinical studies have shown that individuals with significant vision loss are at heightened risk of accidents, particularly during activities involving motion and orientation. Recent research on assistive devices has leveraged deep learning techniques to enhance safety, particularly by detecting stairs and preventing falls. This project aims to evaluate and compare the performance of well-known convolutional neural networks (CNNs) in detecting and classifying stairs, specifically MobileNetV2, GoogleNet, and AlexNet. A dataset of 3,000 RGB images, captured at a resolution of 2268 x 4032 pixels, was used, featuring images of stairs (upstairs, downstairs) and non-stair elements. The labeled dataset was augmented to match the input layer size of the pre-trained models and processed using MATLAB R2022b. Model performance was assessed by analyzing training and validation accuracy and loss through a training progress graph. Additionally, testing accuracy, precision, recall, and F1 score were evaluated using a confusion matrix. The results demonstrated that AlexNet outperformed GoogleNet and MobileNetV2, achieving an impressive 99% accuracy across all performance metrics.
References
Department of Ophthalmology, The University of Pittsburgh. 2022. What is Vision Impairment? Retrieved November 20, 2022, from http://ophthalmology.pitt.edu/vision-impairment/what-vision-impairment
World Health Organization. 2022. Vision Impairment and blindness. World Health Organization. Retrieved November 20, 2022, from https://www.who.int/news room/fact sheets/detail/blindness and visual-impairment
Elliott, D. B., Foster, R. J., & Whitaker, D. 2015. Falls and stair negotiation in older people and their relationship with vision. In Analysis of lower limb movement to determine the effect of manipulating the appearance of stairs to improve safety: A linked series of laboratory-based, repeated measures studies 8th ed., 3. Essay. NIHR Journals Library
Close J, Ellis M, Hooper R, Glucksman E, Jackson S, Swift C. 1999. Prevention of falls in the elderly trial (PROFET): a randomised controlled trial. Lancet. 353: 93–7. DOI: https://doi.org/10.1016/S0140-6736(98)06119-4.
Jack CI, Smith T, Neoh C, Lye M, McGalliard JN. 1995. Prevalence of low vision in elderly patients admitted to an acute geriatric unit in Liverpool: elderly people who fall are more likely to have low vision. Gerontology. 41: 280–5. DOI: https://doi.org/ 10.1159/000213695.
Sharma, H., Tripathi, M., Kumar, A., & Gaur, M. S. 2018. Embedded Assistive Stick for visually impaired persons. 2018 9th International Conference on Computing, Communication and Networking Technologies (ICCCNT). DOI: https://doi.org/10.1109/icccnt.2018.8493707
Matsumura, H., & Premachandra, C. 2022. Deep-learning-based stair detection using 3D point cloud data for preventing visually impairing walking accidents. IEEE Access, 10: 56249–56255. DOI: https://doi.org/10.1109/access.2022.3178154
C. Wang, Z. Pei, S. Qiu, and Z. Tang, 2022. "RGB-D-based stair detection using deep learning for autonomous stair climbing," arXiv preprint arXiv:2212.01098,
Panchi, N., Agrawal, K., Patil, U., Gujarathi, A., Jain, A., Namdeo, H., & Chiddarwar, S. S. 2019. Deep learning-based stair segmentation and behavioral cloning for autonomous stair climbing. International Journal of Semantic Computing, 13(04): 497–512. DOI: https://doi.org/10.1142/s1793351x1940021x
Patil, U., Gujarathi, A., Kulkarni, A., Jain, A., Malke, L., Tekade, R., Paigwar, K., & Chaturvedi, P. 2019. Deep learning-based stair detection and statistical image filtering for autonomous stair climbing. 2019 Third IEEE International Conference on Robotic Computing (IRC). DOI: https://doi.org/10.1109/irc.2019.00031
Ramalingam, B., Elara Mohan, R., Balakrishnan, S., Elangovan, K., Félix Gómez, B., Pathmakumar, T., Devarassu, M., Mohan Rayaguru, M., & Baskar, C. 2021. Stetro-deep learning powered staircase cleaning and maintenance reconfigurable robot. Sensors, 21(18): 6279. DOI: https://doi.org/10.3390/s21186279
N. Rekhawar, Y. Govindani, and N. Rao, 2022. "Deep learning-based detection, segmentation and vision-based pose estimation of staircase," in 2022 1st International Conference on the Paradigm Shifts in Communication, Embedded Systems, Machine Learning and Signal Processing (PCEMS), 78-83. IEEE.
Russell, S. J., & Norvig, P. 2021. Artificial intelligence: A modern approach (4th ed.). Pearson
Albawi, S., Mohammed, T. A., & Al-Zawi, S. 2017. Understanding of a convolutional neural network. 2017 International Conference on Engineering and Technology (ICET). DOI: https://doi.org/10.1109/icengtechnol.2017.8308186
Kumar, A. 2022. Different types of CNN Architectures explained: Examples. Data Analytics. Retrieved December 11, 2022, from https://vitalflux.com/different-types-of-cnn-architectures-explained-examples/
Yamashita, R., Nishio, M., Do, R. K., & Togashi, K. 2018. Convolutional Neural Networks: An overview and application in Radiology. Insights into Imaging, 9(4), 611–629. DOI: https://doi.org/10.1007/s13244-018-0639-9
