Region Estimation System for Manufacturing Joint Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing region estimation systems for manufacturing sites require workers to wear terminals, increasing costs and reducing efficiency, and necessitate new learning when adding joints or regions to be estimated.
Innovation Solution
A region estimation system that detects joint points from images and estimates regions based on these points, allowing for the correction of detected joint points and acquisition of color information without the need for worker-worn terminals or new learning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If worker terminals are used to correct joint points in blind spots, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent uses a pre-trained machine learning model to copy and generate estimated joint points for blind spot regions based on visible joint points and image data. This virtual copying approach eliminates the need for physical worker terminals while maintaining estimation accuracy.
Solution Approach 2:
The patent introduces an intermediary processing layer that uses pre-trained models and virtual joint points to bridge the gap between camera observations and complete body pose estimation. This intermediary system corrects blind spot issues without requiring direct worker input.
2Measurement precision
If worker terminals are required for accurate estimation, then measurement precision is improved, but productivity decreases
Solution Approach 1:
The system performs self-service by automatically estimating and correcting joint points using pre-trained machine learning models and virtual joint points. This eliminates the need for workers to manually input data or wear additional terminals, thereby maintaining high productivity while ensuring accurate measurement.
3Adaptability or versatility
If new joints or regions are added to estimation, then adaptability is improved, but device complexity increases due to new learning requirements
Solution Approach 1:
The patent employs a universal pre-trained machine learning model that can estimate multiple joint points and body regions through a single integrated system. This multi-functional model handles various estimation tasks without requiring separate learning processes for each joint or region, thereby improving adaptability while controlling complexity.
Data Source
AI summary
A region estimation system is provided which can complement a joint or a region while eliminating a need for a worker to wear a terminal and a need to perform new learning is not required even in a case where the joint or the region desired to be estimated is not estimated for some reason. A region estimation system includes: a detector that detects a joint point of an object from an image of the object; and an estimator that estimates a region to be estimated, based on the joint point detected by the detector.