Técnicas de visión artificial en la identificación de objetos para personas no videntes.
This thesis proposes the development of a portable and low-cost artificial vision system aimed at improving the autonomy and quality of life of visually impaired individuals. The ESP32 microcontroller was selected due to its low energy consumption (0.3 W) and affordability, combined with the OV2640...
Saved in:
| Main Author: | |
|---|---|
| Format: | bachelorThesis |
| Language: | deu |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://repositorio.uteq.edu.ec/handle/43000/9133 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Add Tag | Summary: | This thesis proposes the development of a portable and low-cost artificial vision system aimed at improving the autonomy and quality of life of visually impaired individuals. The ESP32 microcontroller was selected due to its low energy consumption (0.3 W) and affordability, combined with the OV2640 camera and the YOLOv3-Tiny algorithm, optimized for resource-constrained hardware. This system enables precise real-time object identification with an average accuracy of 87% and an approximate latency of 500 ms. Additionally, multimodal feedback (auditory and tactile) was determined to be the most effective method for conveying information to users, as it combines detailed descriptions with immediate alerts about nearby obstacles. User testing demonstrated that this modality significantly enhances spatial understanding and reduces cognitive load. The proposed system is not only technically feasible but also economically accessible, making it a scalable solution for developing countries such as Ecuador. The results highlight the potential of integrating artificial vision into portable devices to assist visually impaired individuals, providing an innovative tool that fosters social inclusion and technological accessibility. The project aligns with the Sustainable Development Goals by promoting inclusion through accessible technology. Future improvements include integrating LiDAR sensors and higher-resolution cameras for complex environments. This solution represents a step forward in fostering independence for 285 million visually impaired people worldwide, demonstrating that technical innovation and accessibility can coexist. |
|---|