Hand-Object Detection for Accurate Manufacturing Work State Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for managing work states in product manufacturing lack accuracy, hindering efficient management of worker activities during manufacturing processes.
Innovation Solution
A work management device comprising a first storage unit for storing a learned model that detects objects in images of workers' hands, a detection unit to identify these objects, and a determination unit to automatically determine work states based on the detected objects, utilizing machine learning techniques such as SSD or YOLO for object detection and state transition models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automatic work state determination is implemented using video analysis, then work management efficiency is improved, but measurement precision of work state is insufficient
Solution Approach 1:
The patent segments the work state determination into multiple independent detection stages: hand region detection, object detection within hand region, tool detection, and work state determination. Each stage uses specialized detection models trained for specific tasks, thereby improving overall measurement precision while maintaining automation efficiency
Solution Approach 2:
The patent performs preliminary detection of hand regions before detecting objects within those regions. By first identifying and isolating hand regions using a hand detection model, the system prepares the data structure needed for subsequent precise object detection, thereby improving work state determination accuracy without compromising management efficiency
Data Source
AI summary
In a work management device, a first storage unit stores a first learned model that outputs a plurality of objects defining each of a plurality of work states forming one process of manufacturing work with respect to a determination target image including a hand image that is an image of a hand of a worker who is performing a product manufacturing work, a detection unit detects the plurality of objects with respect to the determination target image by using the first learned model, and a determination unit determines a work state indicated by the determination target image based on the plurality of objects detected by the detection unit.