Silhouette Image Generation for Accurate Shape Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in accurately calculating measurement data for complex shapes of objects, particularly when the shape of the target object is not identified, leading to inaccuracies in product manufacturing processes.
Innovation Solution
A measurement data calculation apparatus that uses a combination of image processing techniques, such as semantic segmentation and GrabCut algorithms, along with dimensionality reduction methods like principal component analysis, to extract and calculate shape data from multiple image angles, enabling precise measurement data generation for product manufacturing systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image processing methods are used to obtain shape information, then the manufacturing process can be simplified, but the measurement accuracy is insufficient for complex shapes
Solution Approach 1:
The patent applies semantic segmentation to divide the image into multiple semantic regions (e.g., object, background, different object parts) and then applies GrabCut algorithm to further segment the object boundary. This multi-stage segmentation approach enables precise extraction of complex shape boundaries while maintaining systematic and manageable processing steps.
Solution Approach 2:
The patent introduces intermediate representations including depth maps, normal maps, and silhouette images as intermediary data structures between the input image and final measurement results. These intermediaries facilitate the transformation process and enable accurate shape extraction by providing additional geometric information at each processing stage.
2Measurement precision
If multiple image processing techniques are combined to improve accuracy, then measurement precision increases, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by first generating depth maps and normal maps from the input image before applying the main segmentation algorithms. These preliminary processed images serve as enhanced inputs that accelerate subsequent processing steps and improve convergence speed of iterative algorithms like GrabCut.
Solution Approach 2:
The patent implements a continuous processing pipeline where each processing stage feeds directly into the next without interruption: image input → depth map generation → normal map generation → semantic segmentation → GrabCut refinement → silhouette extraction → measurement calculation. This continuous flow minimizes idle time and maintains processing efficiency throughout the multi-stage workflow.
Data Source
AI summary
It provides the measurement data of each portion of the target object with a high accuracy. The measurement data calculation apparatus 1020 includes an obtaining unit 1024A, an extraction unit 1024B, a conversion unit 1024C, and a calculation unit 1024D. The obtaining unit 1024A obtains the image data in which the target object is photographed and the full length data of the target object. The extraction unit 1024B extracts the shape data indicating the shape of the target object from the image data. The conversion unit 1024C converts and silhouettes the shape data based on the full length data. The calculation unit 1024D uses the shape data converted by the conversion unit 1024C to calculate the measurement data of each portion of the target object.