Occupancy Level Calculation Using Probability Distributions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for creating obstacle maps using distance sensors face challenges in accurately setting occupancy levels, especially when sensor accuracy is low, leading to difficulties in determining appropriate occupancy values.
Innovation Solution
An information processing apparatus and method that derive an occupancy level by calculating a weighted sum of existence and non-measurement probability distributions based on sensor accuracy and position information, allowing for accurate occupancy level determination even with low sensor accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor accuracy is low, then measurement capability is reduced, but occupancy level determination becomes unreliable
Solution Approach 1:
The patent transforms the occupancy level determination from a single-value assignment to a probability distribution representation. By changing the parameter from a deterministic occupancy level to a probabilistic distribution (existence probability and non-measurement probability), the system can reliably represent uncertainty when sensor accuracy is low, thus maintaining reliability despite reduced measurement precision.
Solution Approach 2:
The patent introduces probability distributions as an intermediary between the sensor measurements and the occupancy level determination. The existence probability distribution and non-measurement probability distribution act as mediators that process the uncertain sensor data, allowing the system to handle low-accuracy measurements while maintaining reliable occupancy representation through weighted summation of these probability distributions.
2Reliability
If occupancy level is set to median value 0.5 for unmeasured regions, then unknown occupancy is represented, but accuracy of obstacle detection is reduced
Solution Approach 1:
The patent changes the occupancy representation from a single median value (0.5) to a probability distribution that captures both existence probability and non-measurement probability. This parameter transformation allows the system to maintain reliable occupancy representation for unmeasured regions while preserving obstacle detection accuracy by distinguishing between different types of uncertainty through the distribution shape.
Solution Approach 2:
The patent segments the occupancy level into two distinct probability components: existence probability (whether an obstacle exists) and non-measurement probability (whether the region was measured). This segmentation allows the system to represent unmeasured regions with appropriate uncertainty while maintaining the ability to accurately detect obstacles in measured regions, thus resolving the contradiction between reliable representation and detection accuracy.
3Measurement precision
If high value of occupancy level is set in vicinity of acquired position, then obstacle presence is indicated, but false positives increase with low sensor accuracy
Solution Approach 1:
The patent introduces probability distributions as an intermediary layer between the sensor acquisition and the final occupancy level setting. The existence probability distribution acts as a mediator that modulates the occupancy level assignment based on both the sensor measurement quality and the spatial relationship to the acquired position. This intermediary processing reduces false positives by only assigning high occupancy levels when the existence probability is sufficiently high, even when sensor accuracy is low.
Solution Approach 2:
The patent changes the occupancy level assignment from a deterministic high value near acquired positions to a probability-based assignment. By transforming the occupancy parameter to include existence probability, the system can indicate obstacle presence near acquired positions while controlling false positives through the probability threshold - only regions with high enough existence probability receive high occupancy levels, maintaining both position indication accuracy and reliability.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
According to an arrangement, an information processing apparatus (20) includes a position acquiring unit (20C), and an occupancy level calculating unit (20G). The position acquiring unit (20C) is configured to acquire position information representing a position where a target exists or no target exists, the position being measured by a sensor (10B). The occupancy level calculating unit (20G) is configured to calculate an occupancy level distribution representing a level of occupancy, by the target, of each of a plurality of regions existing along a direction from the position of the sensor (10B) to the position indicated by the position information based on the position information and measurement accuracy of the sensor (10B), the occupancy level distribution being based on a non-measurement probability distribution representing a probability that measurement is not performed by the sensor.