Dual-AI Object Detection for Navigation Around Occluded Hazards
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional electronic apparatuses fail to identify objects that are partially blocked by other objects, leading to potential collisions and damage during navigation.
Innovation Solution
An electronic apparatus equipped with a camera, sensor, and dual artificial intelligence models, where one model identifies visible objects and another model identifies partially hidden dangerous objects, allowing the apparatus to determine distances and adjust its travel path to avoid collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single AI model is used to identify objects in the image, then the device complexity is reduced, but the measurement precision of partially blocked objects deteriorates
Solution Approach 1:
The patent divides the object identification task into two separate AI models: a first AI model for identifying completely visible objects and a second AI model for identifying partially blocked objects. This segmentation allows each model to be specialized for its specific task, improving overall identification accuracy while managing system complexity through functional division.
2Device complexity
If the electronic apparatus uses conventional object identification methods, then the device complexity is low, but the reliability of collision avoidance deteriorates when objects are partially blocked
Solution Approach 1:
The patent implements preliminary classification of objects based on their visibility status before proceeding with identification. The system first determines whether objects are completely visible or partially blocked, then routes them to appropriate AI models. This preliminary action ensures that partially blocked objects receive specialized processing, improving collision avoidance reliability.
3Productivity
If the electronic apparatus processes only visible objects, then the processing speed is maintained, but the loss of information about blocked objects increases
Solution Approach 1:
The patent introduces an intermediary classification mechanism that identifies partially blocked objects and routes them to a specialized second AI model. This intermediary layer ensures that blocked objects are not overlooked while maintaining efficient processing through targeted model selection, reducing information loss without significantly impacting overall processing speed.
Data Source
AI summary
Provided is an electronic apparatus, including: a sensor; a camera; a storage configured to store first and second artificial intelligence models; a first processor; and a second processor. The second processor is configured to: input an image obtained through the camera to the first artificial intelligence model and identify the object of the first type in the image, and input the image obtained through the camera to the second artificial intelligence model and identify the object of the second type, at least a part of which is not viewable in the image. The first processor is configured to: determine a distance between the object of the first type and the object of the second type based on sensing data received from the sensor; and control the electronic apparatus to travel based on the determined distance.