Surface Scan Comparison for Real-Time Pose Estimation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current pose estimation methods, such as the spin-image algorithm, are computationally complex and impractical for real-time applications due to the time-consuming process of comparing large numbers of spin-images, which is necessary for accurate surface data acquisition and assessment, especially in fields requiring rapid and precise object identification and comparison.

Innovation Solution

A surface data acquisition, storage, and assessment system that reduces the number of comparisons by representing spin-images with fewer numbers, using a similarity measure to group matches, and employing advanced computing hardware like GPUs for parallel processing, allowing for quicker pose estimation and real-time comparisons.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the spin-image algorithm is used for pose estimation, then measurement precision is improved, but productivity deteriorates due to computational complexity

Engineering Contradiction:

Improvepose estimation accuracyVSAvoidreal-time processing speed

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent segments the spin-image comparison process into distinct phases: feature extraction, spin-image generation, and comparison. By identifying and optimizing critical paths within these segments, the system achieves faster processing while maintaining accuracy. The segmentation allows parallel processing of independent spin-images using GPU architecture.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent replaces traditional CPU-based sequential processing with GPU-based parallel processing. This substitution leverages the massively parallel architecture of GPUs to perform simultaneous spin-image comparisons, dramatically increasing processing speed while maintaining the mathematical rigor of the spin-image algorithm for accurate pose estimation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If the number of spin-image comparisons is increased, then measurement precision is improved, but loss of time increases

Engineering Contradiction:

Improvesurface data assessment accuracyVSAvoidcomparison processing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary actions by pre-computing and storing spin-images for reference surfaces in a database. During assessment, only comparison operations are needed rather than full regeneration of spin-images. This preliminary preparation significantly reduces the time required for actual comparisons while enabling comprehensive accuracy through multiple comparisons.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent implements continuous processing by pipelines the data flow from surface scanning through spin-image generation to comparison and assessment. The GPU architecture maintains continuous utilization by processing multiple spin-images in parallel without idle time, ensuring that the increased number of comparisons does not proportionally increase total processing time.

Inventive Principle:

Principle #20Continuity of useful action

Data Source

PatentUS11281176B2Surface data acquisition, storage, and assessment system

Publication Date: 2022.03.22 ECTOSCAN SYSTEMS LLC

AI summary

A surface data acquisition, storage, and assessment system for detecting and quantifying similarities or differences between stored data and data collected from a scan. The system operates to scan features of an object and weighs various parameters and utilizing fuzzy logic makes a recommendation.