Object Recognition Correlation Using Density-Adaptive Matching Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object recognition devices face challenges in accurately calculating correlations between prediction data and detection data due to potential erroneous or non-correlations, especially when multiple detection data points are close or absent, leading to inaccurate object recognition.
Innovation Solution
An object recognition device and method that adjust the determination region based on detection data density, using a time measurement unit, data receiving unit, prediction processing unit, and correlation processing unit to generate adjusted determination region parameters, ensuring accurate correlation between prediction and detection data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed correlation range is used for determining correlation between prediction data and detection data, then the device complexity is reduced, but measurement precision deteriorates due to erroneous correlation or non-correlation
Solution Approach 1:
The patent applies dynamics by making the determination region size variable rather than fixed. The correlation determination unit dynamically adjusts the determination region size based on the detection data density in the prediction region. When detection data density is high, the determination region is reduced to avoid erroneous correlation; when density is low, the region is expanded to prevent non-correlation. This dynamic adjustment resolves the contradiction between simple fixed-range implementation and accurate correlation determination.
Solution Approach 2:
The patent changes the parameter of determination region size based on detection data density. By calculating the density of detection data within the prediction region and using this density value to adjust the determination region size, the system adapts the correlation range to actual data conditions. This parameter change enables accurate correlation determination without requiring complex fixed-range management, resolving the contradiction between simplicity and precision.
2Measurement precision
If the determination region size is reduced to prevent erroneous correlation, then correlation accuracy is improved, but non-correlation occurs when detection data is sparse
Solution Approach 1:
The system dynamically adjusts the determination region size based on real-time detection data density. When detection data is dense, the region shrinks to prevent erroneous correlation; when data is sparse, the region expands to ensure detection data is captured. This dynamic behavior simultaneously improves correlation accuracy and maintains reliability across varying data conditions.
Solution Approach 2:
The determination region size parameter is changed according to detection data density calculations. By using the density value as a basis for parameter adjustment, the system ensures that the region is appropriately sized for current data conditions, preventing both erroneous correlation (when too large) and non-correlation (when too small), thus resolving the contradiction between accuracy and reliability.
3Reliability
If the determination region size is increased to ensure detection data coverage, then correlation reliability is improved, but erroneous correlation occurs when multiple detection data points exist
Solution Approach 1:
The system changes the determination region size parameter based on detection data density. When multiple detection data points are present (high density), the region is reduced to prevent erroneous correlation while still capturing the correct data point. When data is sparse (low density), the region is increased to ensure coverage. This adaptive parameter change resolves the contradiction between reliability and accuracy.
Data Source
AI summary
An object recognition device includes a time measurement unit to measure a time, a data receiving unit to receive detection data from a plurality of sensors and associate the time with the detection data, a received data processing unit to calculate a detection data density, a prediction processing unit to predict a state value of the object corresponding to the associated time from a state value of the object at an immediately preceding associated time and generate a prediction result as prediction data, an adjusted determination region parameter generation unit to generate an adjusted determination region parameter by adjusting a parameter indicating a size of a determination region, a correlation processing unit to generate correlation data between the prediction data and the detection data, and an update processing unit to update the state value of the object on the basis of the correlation data.