Spatial Accumulator Merging Point Clouds for 3D Profilometry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current three-dimensional inspection systems for circuit boards are inefficient in high-speed automated inspection due to slow processing times and inability to effectively merge point clouds from multiple cameras, especially when dealing with shiny objects and partial occlusions, leading to errors in solder paste deposition and component placement.
Innovation Solution
A computer-implemented method and system that projects structured illumination and acquires multiple sets of images to generate point clouds, which are then combined using a spatial accumulator to reduce errors from multiple reflections and occlusions, employing a spatial accumulator to merge complex values from multiple cameras and apply non-linear techniques like geometric median for robust estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple cameras are used to capture point clouds for three-dimensional inspection, then measurement precision and detection capability are improved, but processing time increases and productivity decreases
Solution Approach 1:
The patent divides the point cloud processing task into multiple independent accumulators, each handling data from specific cameras. This segmentation allows parallel processing of point clouds from different camera views, maintaining high measurement precision while reducing overall processing time through concurrent operations.
Solution Approach 2:
The patent performs preliminary actions by pre-defining accumulator structures and preparing data routing paths before point cloud acquisition. This pre-configuration enables immediate parallel processing upon data arrival, eliminating setup delays and maintaining high inspection speed while handling multiple camera inputs.
2Reliability
If traditional point cloud merging tools like PCL are used, then comprehensive processing capability is achieved, but processing speed becomes too slow for high-speed automated inspection
Solution Approach 1:
The patent segments the monolithic point cloud merging process into multiple independent accumulators that operate in parallel. Each accumulator handles a specific subset of camera data with dedicated processing logic, providing comprehensive processing capability while achieving the speed necessary for high-speed automated inspection through concurrent execution.
Solution Approach 2:
The patent introduces accumulators as intermediary data structures that receive and process point cloud data from multiple cameras before final merging. These accumulators act as mediators that simplify the complex merging operation into manageable stages, ensuring processing completeness while maintaining high speed through efficient intermediate processing.
3Area of stationary object
If point clouds from multiple cameras are merged using conventional methods, then complete surface coverage is achieved, but errors from multiple reflections and occlusions increase
Solution Approach 1:
The patent segments the point cloud data by creating separate accumulators for different camera views and spatial regions. This segmentation allows independent error analysis and correction for each accumulator, maintaining complete surface coverage while reducing the impact of reflection and occlusion errors through localized processing and validation.
Solution Approach 2:
The patent implements feedback mechanisms within each accumulator to detect and correct errors from multiple reflections and occlusions. By continuously monitoring point cloud quality metrics and adjusting processing parameters based on detected errors, the system maintains high measurement precision across the complete surface area while compensating for optical interference.
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
This approach significantly reduces point cloud errors, enhances accuracy in detecting defects such as lifted leads and component misplacement, and improves the speed of inspection, making it suitable for high-speed automated inspection systems in the electronics industry.
Implementation Method 1
projecting structured illumination on the surface and acquiring a plurality of sets of images
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer-implemented method (100) of and system (60) for measuring a three- dimensional surface are provided. The method (100) includes projecting (102) structured illumination on the surface and acquiring a plurality of sets of images. The sets of images are processed (106) to obtain a plurality of point clouds. A spatial accumulator is defined (104). A first point cloud of the plurality of point clouds is combined (108) with a second point cloud of the plurality of point clouds into the spatial accumulator. Spatial coordinates of the surface are generated (112) based on the contents of the spatial accumulator.