Illumination Panel Imaging for Accurate Target Mask Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in accurately distinguishing between targets and backgrounds in captured images, leading to low accuracy in extracting target images, particularly when the luminance of the background and target are similar, and manual corrections are often necessary.
Innovation Solution
A data obtaining system that uses an illumination panel to emit light and a capture device to capture images, generating mask data based on the difference between images with and without a target, thereby increasing contrast and enabling accurate extraction of target images without manual corrections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional imaging without specialized illumination is used, then the system is simple, but the accuracy of target image extraction is low when target and background luminance are similar
Solution Approach 1:
The illumination panel emits light in specific colors (e.g., red, green, blue) to create contrast between the target and background. By changing the color of emitted light, the system enhances the distinguishability of the target from the background, thereby improving extraction accuracy without requiring complex mechanical or structural modifications.
Solution Approach 2:
The system performs preliminary actions by capturing a background image without the target first, then capturing an image with the target. Mask data is generated by comparing these two images, which preliminarily separates the target from the background before final extraction. This preliminary differentiation step significantly improves extraction accuracy.
2Measurement precision
If manual corrections are performed to improve extraction accuracy, then extraction precision improves, but productivity decreases due to time-consuming manual intervention
Solution Approach 1:
The system performs self-service by automatically generating mask data through image comparison and processing. The controller automatically captures images, compares them to generate mask data, and extracts target images without requiring manual intervention. This automation maintains high extraction accuracy while significantly improving productivity by eliminating time-consuming manual corrections.
Solution Approach 2:
The system replaces manual mechanical correction processes with automated image processing algorithms. Instead of manually adjusting and correcting target image extraction, the system uses computational methods to automatically generate mask data and extract targets, substituting human labor with automated processing that maintains accuracy while improving efficiency.
3Measurement precision
If the illumination panel emits light in multiple colors, then the accuracy of distinguishing target from background improves, but the complexity of controlling the illumination panel increases
Solution Approach 1:
The illumination panel is divided into multiple regions or zones that can emit different colors of light independently. By segmenting the illumination panel, the system can emit different colors in different areas to enhance target-background distinction. This segmentation approach improves distinction accuracy while keeping control manageable by treating each segment as an independent controllable unit.
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 system enhances the accuracy of target image extraction by increasing contrast between the illumination panel and the target, allowing for simplified annotations and improved training data generation for recognition models.
Implementation Method 1
The controller causes the illumination panel to emit light
Data Source
AI summary
A data obtaining device includes a controller capable of controlling an illumination panel and capable of obtaining at least one captured image, which is obtained by capturing at least one image of an illumination surface of the illumination panel. The controller generates mask data for a target located in front of the illumination panel on a basis of, among the at least one captured image, a captured image of the illumination panel and a target with the illumination panel emitting light.