Dual Imaging Position Detection for Overlapping Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing position detection systems face challenges in accurately determining the position of objects, particularly when objects are hidden or overlapping, leading to inaccuracies and misidentification.
Innovation Solution
A system comprising two imaging units and a controller that calculates the position of an object by defining position and direction vectors based on captured images, with the controller distinguishing between 'mask state' and 'cascade state' to handle object invisibility or overlap, improving detection precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single imaging unit is used to detect object position, then the system structure is simple, but detection accuracy deteriorates when objects are hidden or overlapping
Solution Approach 1:
The system divides the detection task into multiple independent imaging units, each capturing images from different spatial positions. This segmentation allows the system to overcome occlusion by having multiple viewpoints, where at least one viewpoint can see hidden objects. The position detection is then performed by integrating data from these segmented imaging units.
2Measurement precision
If multiple imaging units are used to improve detection accuracy, then object position detection accuracy improves, but device complexity increases
Solution Approach 1:
The system merges the detection results from multiple imaging units through a coordinated processing mechanism. The controller integrates images and detection data from all imaging units, combining their individual detection results to achieve accurate object position detection even when objects are hidden or overlapping, thereby improving overall system accuracy without requiring each unit to be independently complex.
3Device complexity
If traditional position detection methods are used, then the detection process is simple, but reliability deteriorates when objects are hidden or overlapping
Solution Approach 1:
The system introduces an intermediary controller that coordinates between multiple imaging units and the detection algorithm. This intermediary component manages the integration of images from different viewpoints, applies the five-state detection algorithm to distinguish between different occlusion scenarios, and synthesizes reliable position information even when individual imaging units cannot detect objects due to hiding or overlapping.
4Speed
If basic image capture is performed without sophisticated processing, then processing speed is fast, but detection reliability worsens when objects are hidden or overlapping
Solution Approach 1:
The system performs preliminary actions by capturing images from multiple imaging units simultaneously before conducting the detection analysis. This preliminary multi-viewpoint image acquisition ensures that even if objects are hidden from one viewpoint, they are captured from other viewpoints. The five-state detection algorithm then processes these pre-captured images to reliably determine object positions without requiring complex real-time processing during the detection phase.
Data Source
AI summary
Provided is a position detection method using a first imaging unit and a second imaging unit that each capture an image of an object. The detection method may include: acquiring a first captured image with use of the first imaging unit; acquiring a second captured image with use of the second imaging unit; defining a position vector of the object based on the first captured image; calculating a first direction vector directed from the second imaging unit to the object with use of the defined position vector; calculating a second direction vector directed from the second imaging unit to the object based on the second captured image; and calculating a position of the object based on the first direction vector and the second direction vector.