3D Point Cloud Filtering With Adaptive Coefficient Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional VSLAM techniques output point group information with significant noise, and it is challenging to set filter coefficients appropriately due to varying conditions in image data captured by cameras, such as sunshine and imaging scenes.
Innovation Solution
An information processing device that uses a filter coefficient estimation model trained with test data and observation data to dynamically set filter coefficients for filters, reducing noise in point group information using a neural network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If various filters are applied to reduce noise in point group information, then noise reduction is achieved, but it is difficult to appropriately set filter coefficients because image data conditions vary significantly
Solution Approach 1:
The system automatically determines filter coefficients through a determination unit that analyzes image data characteristics and selects appropriate coefficients without manual intervention. This self-service mechanism resolves the contradiction by making the system autonomously adapt to varying image conditions, eliminating the operational difficulty while maintaining noise reduction effectiveness
Solution Approach 2:
The filter coefficients are dynamically adjusted based on the characteristics of the input image data. The determination unit evaluates current image conditions and modifies coefficients in real-time, transforming the static filtering process into a dynamic adaptive system that maintains optimal performance across varying conditions
2Device complexity
If manual filter coefficient setting is used, then device complexity is reduced, but measurement precision deteriorates due to inability to adapt to varying image conditions
Solution Approach 1:
A determination unit is introduced as an intermediary component between the image data input and the filtering process. This intermediary automatically analyzes image characteristics and selects appropriate filter coefficients, adding computational complexity but enabling adaptive precision that manual setting cannot achieve
Solution Approach 2:
The system automatically changes filter coefficients based on detected parameters of the input image data. By dynamically adjusting these parameters according to image characteristics such as lighting conditions and scene complexity, the system maintains high measurement precision without requiring manual intervention
Data Source
AI summary
An information processing device includes a memory and a processor configured to obtain first point group information that represents three-dimensional positional information, by using image data captured by a first camera; output second point group information obtained by reducing noise of the first point group information, by using one or more filters; and set filter coefficients of the one or more filters, by using (i) a filter coefficient estimation model trained in advance by using (a) test data for learning that includes image data captured by a second camera and (b) training data based on third point group information that represents three-dimensional positional information obtained by a position sensor, and (ii) observation data that includes the image data captured by the first camera.