Dual Detector Object Detection with Dynamic Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object detection techniques face challenges in maintaining detection performance, particularly when the training of correspondence relationships does not converge, leading to erroneous or unstable detections due to factors like weather fluctuations and object orientation, especially in outdoor environments.
Innovation Solution
An information processing device that uses two detectors, where the detection result of a first detector is used to determine the settings for a second detector, allowing for improved detection accuracy by adjusting parameters such as threshold values and calculation resources based on prior detection information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single detector is used for object detection, then the device complexity is reduced, but the detection stability and accuracy deteriorate when training does not converge
Solution Approach 1:
The detection system is segmented into multiple detectors, each responsible for detecting different targets. The first detector detects a first target while the second detector detects a second target. This segmentation allows each detector to be optimized for specific detection tasks, improving overall detection stability without requiring a single complex detector to handle all cases
Solution Approach 2:
The detection system is designed with multi-functionality where detectors can be configured to detect different targets based on detection history. The system universally handles various detection scenarios by switching between different detector configurations and processing modes, maintaining reliability across diverse detection requirements
2Ease of operation
If threshold values are fixed for object detection, then the ease of operation is improved, but the detection accuracy deteriorates under varying conditions
Solution Approach 1:
The threshold values are made dynamic rather than fixed. The processing parameter for the second detector is determined based on the detection history from the first detector. This dynamic adjustment allows the system to adapt thresholds to varying detection conditions and target states, maintaining high detection accuracy without complex manual configuration
Solution Approach 2:
The system implements feedback mechanisms where the detection result from the first detector informs the configuration of the second detector. The processing parameter of the second detector is determined based on the detection history, creating a feedback loop that continuously optimizes detection accuracy based on actual detection conditions and historical performance
3Measurement precision
If detection parameters are adjusted based on detection history, then the detection accuracy is improved, but the processing time increases
Solution Approach 1:
The system performs preliminary detection actions using the first detector to gather detection history information before configuring the second detector. By conducting initial detection and using those results to inform subsequent detection parameter settings, the system prepares optimal detection configurations in advance, reducing overall processing time while maintaining accuracy
Solution Approach 2:
The detection system operates in periodic cycles where the first detector processes detection tasks, then the system periodically updates the processing parameters for the second detector based on accumulated detection history. This periodic parameter adjustment balances the need for accurate detection with efficient processing by not continuously recalculating parameters
Data Source
AI summary
An information processing device is an information processing device including a processor. The processor obtains a detection result of a first detector for detecting a first target in first sensing data; and based on the detection result of the first detector, determines a setting of processing by a second detector for detecting a second target in second sensing data next in an order after the first sensing data, the second target being different from the first target.