Arithmetic Processor for Ranging Device Background Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Ranging devices face challenges in stabilizing background detection due to variations in observed data, particularly when objects with subtle movements or complex structures like wire meshes are present, leading to incorrect classification of stationary objects as moving objects.
Innovation Solution
An arithmetic processor is designed to process data from ranging devices, utilizing a background candidate memory, ranking section, observation count update, and background determination section to identify stable backgrounds by correlating observation counts and variance, ensuring accurate background detection even with varying data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the background subtraction technique is applied to ranging device data, then moving objects can be detected, but stationary objects like wire meshes may be incorrectly classified as moving objects due to data variation
Solution Approach 1:
The patent applies preliminary action by performing multiple preliminary observations and ranking background candidates before final determination. The system collects multiple observed data points, ranks them by observation count, and only after sufficient preliminary validation does it determine the background, thereby resolving the contradiction between stability and precision.
Solution Approach 2:
The patent implements feedback through the observation count mechanism. Each background candidate is continuously evaluated based on feedback from multiple observations, and the ranking is updated accordingly. This feedback loop ensures that only consistently observed backgrounds are selected, improving both reliability and precision.
2Measurement precision
If multiple background candidates are ranked and validated over a predetermined period, then accurate background detection is achieved, but processing time increases
Solution Approach 1:
The patent applies partial action by ranking only the top N background candidates and determining backgrounds only when M candidates (where M≤N) consistently rank in the top positions. This partial processing approach maintains high accuracy while reducing the time loss compared to evaluating all possible candidates.
3Measurement precision
If the observation count threshold is set high to ensure accuracy, then background detection precision improves, but the system becomes less responsive to changes
Solution Approach 1:
The patent implements dynamics by allowing the background determination to adapt based on the consistent ranking of M out of N candidates over a predetermined period. This dynamic approach balances precision and responsiveness by not requiring all N candidates to meet the threshold,而是 allowing a proportion M to determine the background, thus maintaining both accuracy and adaptability.
Data Source
AI summary
An arithmetic processor is provided to detect a background even when observed data values vary. The processor includes a background candidate data memory 20 storing data as a background candidate. In the data, an observation count, as the number of times of obtaining the same observation data, is correlated with distance data representing the observed data. The processor also includes a ranking section 11 ranking background candidates using the observation count as a basis; an observation count update section 12 comparing observed data with a background candidate stored in the memory 20 to find a match, counting observations of the background candidate if there is a match, and replacing distance data of a lowest-ranked background candidate with the observed data if there is no match; and a background determination section 13 determining there are two backgrounds if the top two background candidates remain unchanged for a predetermined period.