ViBe Sonar Point Cloud Segmentation for Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Three-dimensional sonar systems face challenges in processing noise points and outliers in point cloud data due to equipment accuracy, environmental factors, and data mosaicking, which affects subsequent processing steps like registration and feature extraction.
Innovation Solution
A ViBe-based method for three-dimensional sonar point cloud image segmentation involves converting sonar data to point cloud data, down-sampling, applying the ViBe algorithm for image segmentation, accumulating scores, clustering, and expanding foreground data to effectively separate noise from valid data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If filtering is performed to remove noise points and outliers, then data validity is improved, but processing time increases and subsequent pipeline processing is affected
Solution Approach 1:
The patent applies preliminary action by performing noise filtering and outlier removal in advance before subsequent processing steps. The ViBe algorithm is used to segment and remove noise points beforehand, ensuring that the data is cleaned prior to registration, feature extraction, and surface reconstruction operations, thus improving data validity without affecting subsequent pipeline processing
Solution Approach 2:
The patent applies segmentation by dividing the point cloud data into multiple segments or clusters. The ViBe algorithm segments the data into foreground (valid points) and background (noise points) based on visual characteristics and spatial distribution, allowing selective removal of noise while preserving valid data structures for subsequent processing
2Reliability
If noise filtering is performed, then data validity is improved, but the complexity of the processing pipeline increases
Solution Approach 1:
The patent applies universality by using the ViBe algorithm to perform multiple functions: noise filtering, outlier removal, and data segmentation all within a single integrated framework. This multi-functional approach consolidates what would otherwise require separate processing steps, reducing overall pipeline complexity while maintaining data validity
Solution Approach 2:
The patent applies the intermediary principle by introducing the ViBe algorithm as a mediator between raw sonar data and subsequent processing steps. The algorithm acts as an intermediate processing layer that cleans and segments the data, simplifying the overall pipeline by providing a standardized interface for noise removal before registration and feature extraction
3Productivity
If point cloud data is processed with noise and outliers, then processing speed is maintained, but image reconstruction quality deteriorates
Solution Approach 1:
The patent applies preliminary action by removing noise and outliers before surface reconstruction and image generation. The ViBe algorithm segments and removes invalid points in advance, ensuring that only clean data is used for reconstruction, thus maintaining both processing speed and reconstruction quality
Solution Approach 2:
The patent applies the extraction principle by extracting and removing noise points and outliers from the point cloud data. The ViBe algorithm identifies and extracts invalid data points based on their visual characteristics and spatial anomalies, leaving only valid foreground points for subsequent reconstruction operations, thereby improving reconstruction quality without significantly impacting processing speed
Data Source
AI summary
The present invention discloses a ViBe-based three-dimensional sonar point cloud image segmentation method, characterized by including: (1) acquiring sonar data, and converting three-dimensional sonar depth image data corresponding to each frame of the sonar data to point cloud data under an orthogonal coordinate system; (2) sampling the point cloud data, and down-sampling the point cloud data to a plurality of adjacent voxels with side length being R by taking a variable resolution R as a function; (3) carrying out image segmentation on the down-sampled point cloud data by a ViBe algorithm; (4) carrying out accumulative scoring on each voxel according to an image segmentation result, and sorting foreground data and background data according to accumulative score; and (5) clustering the foreground data, and then carrying out expansion operation on the foreground data by taking an original point in the voxel as a center to obtain final foreground point cloud data.