Arc Welding Spatter Detection Using Color-Based Reflection Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing welding condition setting assistance devices inaccurately detect spatters due to arc light reflection from peripheral devices, leading to incorrect identification of the number of spatters.
Innovation Solution
A welding condition setting assistance device that includes an image processing unit to detect spatter candidate regions based on pixel brightness, identifies reflected light regions using color information, and removes these regions to accurately count spatters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If spatters are detected based on pixel brightness values in input images, then the detection process is simple and fast, but arc light reflected from peripheral devices is erroneously detected as spatters, reducing measurement precision
Solution Approach 1:
The patent applies color information analysis to distinguish between genuine spatters and reflected arc light. By examining the color characteristics (RGB values) of detected bright regions, the system can identify that reflected light typically has different color properties than actual spatters, thereby eliminating false detections while maintaining detection speed
Solution Approach 2:
The patent introduces color information as an intermediary criterion between simple brightness detection and accurate spatter identification. This intermediate step uses color analysis to filter out false positives from reflected light, resolving the contradiction between detection simplicity and measurement accuracy
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The device accurately identifies the number of spatters by distinguishing between genuine spatters and reflected light, enhancing the precision of spatter detection during arc welding.
Implementation Method 1
detection of a spatter candidate region based on a pixel value indicating brightness of a pixel included in the input images
Implementation Method 2
identifying, in the spatter candidate region detected in the spatter candidate region detection step, a reflected light region in which reflected light of arc light is shown, based on color information of a predetermined reference pixel included in the spatter candidate region
Data Source
AI summary
Provided is image processing unit that causes computer to perform: a spatter candidate region detection step of performing, for each of input images obtained by capturing workpiece during arc welding, detection of a spatter candidate region based on a pixel value indicating brightness of a pixel included in the input images; a reflected light region identification step of identifying, in the spatter candidate region detected in the spatter candidate region detection step, a reflected light region in which reflected light of arc light is shown, based on color information of a reference pixel included in the spatter candidate region; and a spatter number identification step of identifying, as the number of spatters, the number of spatter candidate regions obtained by removing the reflected light region identified in the reflected light region identification step in the spatter candidate region detected in the spatter candidate region detection step.