3D Voxel Imaging for Fast Multi-Camera Shape Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating three-dimensional images of target objects, such as bonding wires, require capturing multiple images by changing the focusing height of the optical system, leading to prolonged processing times.
Innovation Solution
A three-dimensional image generation device utilizing multiple cameras and a control unit to set voxels in a space, image these voxels from different directions, detect brightness levels, specify specific voxels based on threshold values, and connect these voxels to generate a three-dimensional image without altering the focusing height.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images are captured by changing the focusing height of the optical system, then the three-dimensional shape can be detected, but the processing time becomes long
Solution Approach 1:
The patent divides the three-dimensional space into multiple voxels and processes each voxel independently by determining brightness from multiple camera images. This segmentation allows parallel processing of different spatial regions, reducing overall processing time while maintaining three-dimensional shape detection accuracy.
Solution Approach 2:
The patent transitions from a single optical system with variable focus (one-dimensional adjustment) to multiple cameras positioned at different locations (two-dimensional spatial arrangement). This dimensional change enables simultaneous capture of multiple views, eliminating the need for sequential focus adjustment and reducing processing time.
2Adaptability or versatility
If a single optical system with shallow focusing height is used, then the wire image can be captured at different depths, but multiple capture operations are required
Solution Approach 1:
The patent merges the functionality of multiple optical systems with different focus settings into a single system using multiple cameras positioned at different locations. Each camera captures images from its specific viewpoint, and the control unit integrates these images to reconstruct three-dimensional shapes, achieving both depth coverage and improved productivity.
Solution Approach 2:
Instead of using a single optical system that physically adjusts focus, the patent uses multiple cameras as copies of the optical system, each fixed at a specific position. These copies simultaneously capture images from different depths, eliminating the need for mechanical focus adjustment and improving image generation speed.
3Measurement precision
If the focusing height is changed to detect different parts of the target object, then complete three-dimensional information can be obtained, but the operation becomes complex
Solution Approach 1:
The patent replaces the mechanical focus adjustment system with a computational approach using multiple fixed cameras. Instead of physically moving the optical system to different focus positions, the control unit processes images from multiple fixed viewpoints to reconstruct three-dimensional information, simplifying the mechanical operation while maintaining measurement precision.
Solution Approach 2:
The patent changes the parameter of focus adjustment to a parameter of camera positioning. Instead of adjusting focus distance, the system uses multiple cameras fixed at different spatial positions to capture images. The control unit then processes these images to obtain three-dimensional information, transforming a mechanical parameter change into a spatial arrangement solution.
Data Source
AI summary
A three-dimensional image generation device and a three-dimensional image generation method are provided. The three-dimensional image generation device includes multiple cameras using imaging elements, and a control unit that processes respective images captured by the cameras. The control unit sets multiple voxels in a space including a target object; images the multiple voxels from multiple directions; (a) detects each brightness of each imaging clement corresponding to one voxel among the multiple voxels; (b) specifies the lowest brightness among the detected brightnesses as the lowest brightness of the one voxel; (c) specifies, when the specified lowest brightness is at least a prescribed threshold value, the one voxel as a specific voxel including the target object; repeatedly executes the operations of (a) to (c) for all of the multiple voxels;and generates a three-dimensional image of the target object by connecting the multiple specific voxels specified in (c).