Neural network combining visible and thermal images for inferring environmental data of an area of a building
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current environmental control systems in buildings do not effectively utilize both visible and thermal images from neural networks to infer environmental data, limiting their accuracy and reliability in determining conditions such as occupancy and activity.
Innovation Solution
A computing device and method that employs a neural network inference engine using both visible and thermal images from cameras to infer environmental data, such as occupancy and activity, by generating a predictive model through a training engine and executing it to determine set points for controlling appliances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only visible images or only thermal images are used as input to a neural network, then the system complexity is reduced, but the accuracy and reliability of environmental data inference deteriorates
Solution Approach 1:
The patent combines visible images and thermal images as dual inputs to a single neural network, merging two different sensing modalities to achieve more accurate and reliable environmental data inference. This merging approach allows the system to leverage complementary information from both image types, improving occupancy detection and activity recognition accuracy while managing complexity through integrated processing architecture.
2Reliability
If visible and thermal images are processed separately by different neural networks, then the processing pipeline is simpler, but the reliability of determining occupancy and activity deteriorates
Solution Approach 1:
The patent merges the processing of visible and thermal images into a single neural network that accepts both image types as simultaneous inputs. This unified approach improves reliability by allowing the network to cross-validate information from both modalities and make more confident determinations about occupancy and activity, while avoiding the need for complex post-processing integration of separate network outputs.
3Adaptability or versatility
If a single neural network processes both visible and thermal images, then the system architecture is more integrated, but the computational resources and processing time increase
Solution Approach 1:
The patent implements a universal neural network architecture that can process both visible and thermal images through a unified framework. This multi-functional system adapts to different environmental control needs by accepting multiple input types and producing comprehensive environmental data, while the integrated design avoids the redundant processing overhead of multiple separate networks, optimizing energy consumption relative to the versatility provided.
Data Source
AI summary
Method and computing device for inferring via a neural network environmental data of an area of a building based on visible and thermal images of the area. A predictive model generated by a neural network training engine is stored by the computing device. The computing device determines a visible image of an area based on data received from at least one visible imaging camera. The computing device determines a thermal image of the area based on data received from at least one thermal imaging device. The computing device executes a neural network inference engine, using the predictive model for inferring environmental data based on the visible image and the thermal image. The inferred environmental data comprise geometric characteristic(s) of the area, an occupancy of the area, a human activity in the area, temperature value(s) for the area, and luminosity value(s) for the area.