LIDAR Super-Resolution via Cloud Point Duplication and Shift Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
LIDAR systems face limitations in generating high-resolution images due to view shifts and quantum conversion inefficiencies, resulting in ambiguity and reduced detail, especially in representing non-illuminated object sides and multiple return values for the same coordinate location.
Innovation Solution
The system enhances LIDAR data sets by duplicating cloud points within each view, compensates for view shifts, and integrates valid points to generate a super-resolved image, improving resolution by averaging or using masks for valid data across multiple views.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple views are combined to improve resolution, then image detail is enhanced, but view shifts cause ambiguity and reduce accuracy
Solution Approach 1:
The patent applies preliminary action by duplicating cloud points before combining multiple views. This preprocessing step creates redundant data points that can be systematically processed to resolve view shift ambiguities later in the integration process, thereby maintaining both high resolution and accuracy
Solution Approach 2:
The patent uses copying by duplicating cloud points from each view multiple times before integration. This creates multiple copies of the same spatial information that can be distributed across different views during combination, allowing the system to recover accurate position information even when views are shifted relative to each other
2Measurement precision
If cloud points are duplicated to enhance views, then detail representation improves, but data processing complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the enhanced point cloud data into valid and invalid portions during integration. By segmenting the duplicated points and selectively processing only valid ones, the system maintains high detail representation while reducing unnecessary computational complexity from processing redundant invalid data
3Loss of information
If multiple return values are captured for the same coordinate, then completeness improves, but ambiguity increases
Solution Approach 1:
The patent uses feedback mechanisms during the integration process to resolve ambiguities from multiple return values. By iteratively processing duplicated cloud points and using validity criteria to feedback on which points to accept or reject, the system maintains complete information while reducing ambiguity through systematic validation
Data Source
AI summary
Light detection and ranging (LIDAR) imaging systems, method, and computer readable media for generating super-resolved images are described. Super-resolved images are generated by obtaining data sets of cloud points representing multiple views of an object where the views have a view shift, enhancing the views by duplicating cloud points within each of the data sets, compensating for the view shift using the enhanced views, identifying valid cloud points, and generating a super-resolved image of the object by integrating valid cloud points within the compensated, enhanced views.