Distributed Architecture to Integrate Sensor Information: Object Recognition for Smart Cities.

Area of Science:

• Computer Science
• Artificial Intelligence
• Internet of Things

Background:

• Object recognition is crucial for smart city applications like environmental mapping and intelligent vehicle navigation.
• Existing systems often lack robust methods for integrating diverse data sources in intelligent environments.

Purpose of the Study:

• To propose and validate a novel architecture for object recognition in intelligent environments.
• To leverage heterogeneous information integration using an IoT/Industry 4.0 model.

Main Methods:

• Developed an architecture based on the IoT/Industry 4.0 model, utilizing interconnected 'smart resources'.
• Smart resources process local sensor data and offer information as services across edge, fog, and cloud computing layers.
• Implemented a validation system with two smart resources featuring different image sensors.

Main Results:

• The proposed architecture successfully integrates heterogeneously distributed information.
• Information integration increased object recognition certainty by 2-4% in experimental validation.
• Demonstrated the effectiveness of collaborative capabilities among intelligent devices.

Conclusions:

• The integration of information from distributed smart resources significantly improves object recognition accuracy.
• Intelligent environments benefit from devices possessing both intelligence and collaborative capabilities.
• The proposed architecture provides a scalable and effective solution for object recognition in smart cities.