Action Determination Device Distance-Based Parameter Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing action determination devices face challenges in accurately determining actions from captured images due to limitations in feature point extraction and tracking, particularly when the distance of the object from the imaging position varies, leading to reduced accuracy and increased noise.
Innovation Solution
The action determination device includes an extraction unit for feature point extraction, a tracking unit for generating tracking information, and a determination unit that compares accumulated tracking information with registered information, with parameter optimization based on distance from the imaging position to improve accuracy and reduce noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If feature points are extracted from captured images without distance-based parameter optimization, then the processing is simple, but the accuracy of action determination deteriorates when the distance of the object from the imaging position varies
Solution Approach 1:
The patent applies parameter changes by adjusting feature point extraction and tracking parameters according to the distance of the object from the imaging position. When the object is far from the imaging position, the extraction threshold is lowered and the number of extracted feature points is increased to maintain determination accuracy. This resolves the contradiction by adapting parameters to distance conditions, improving accuracy without requiring a completely different processing system.
Solution Approach 2:
The patent implements dynamics by making the feature point extraction and tracking parameters variable rather than fixed. The parameters dynamically adjust based on the detected distance of the object, allowing the system to adapt to different imaging conditions. This dynamic adjustment enables the system to maintain high accuracy across varying distances while using a unified processing framework.
2Measurement precision
If tracking information from multiple frames is accumulated and compared as a group, then the accuracy of action determination improves, but the processing time and computational load increase
Solution Approach 1:
The patent applies preliminary action by pre-registering action patterns and their corresponding tracking information groups before actual action determination. The registered information groups containing multiple frames of tracking data are prepared in advance, allowing for efficient comparison during real-time processing. This pre-preparation reduces the computational burden during actual action recognition.
Solution Approach 2:
The patent uses copying by creating registered information groups that are copies of expected action patterns stored in advance. During action determination, the system compares the current tracking information group against these pre-stored copies of action patterns, enabling efficient pattern matching without requiring complex real-time analysis of each frame individually.
3Reliability
If the extraction threshold is fixed for all distances, then the processing is simple, but noise increases and tracking errors occur when the object distance varies
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the extraction threshold based on the object's distance from the imaging position. When the object is far away, the threshold is lowered to capture more feature points, while for closer objects, a higher threshold is used to reduce noise. This adaptive thresholding improves tracking reliability across different distances without requiring a completely different processing architecture.
Solution Approach 2:
The patent implements local quality by applying different extraction and tracking parameters to different distance conditions. The system identifies the distance of the object and applies locally optimized parameters suited to that specific condition, rather than using a uniform approach. This allows each distance range to be processed with parameters that maximize reliability for that condition.
Data Source
AI summary
An action determination device includes: an extraction unit that extracts feature points of an object from captured images; a tracking unit that generates tracking information indicating a moving direction of the object based on the feature points respectively extracted from the captured images temporally preceding and succeeding each other; and a determination unit that determines whether or not an action is performed based on a comparison result between a tracking information group in which a plurality of pieces of the tracking information are accumulated in time series and a registered information group registered in advance in association with the action of the object.