Non-Perspective Lens 3D Vision System Pose Accuracy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing 3D machine vision systems face challenges in accurately determining the pose of microscopic and near-microscopic objects due to limitations in camera lens arrangements, particularly with small baselines and perspective projections, leading to reduced accuracy and increased processing overhead.

Innovation Solution

The use of non-perspective lenses, such as telecentric lenses, in conjunction with multiple cameras allows for the acquisition of images that are less sensitive to object tilt and position changes, enabling more efficient determination of 3D pose through affine transformations and improved feature correspondence.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If perspective lenses are used in stereo camera systems, then the system can capture images with depth information, but the accuracy of pose determination decreases due to sensitivity to object tilt and position changes

Engineering Contradiction:

Improvepose determination accuracyVSAvoidprocessing overhead

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes the optical parameter of the lens system by using non-perspective (telecentric) lenses instead of conventional perspective lenses. This parameter change eliminates perspective distortion and makes the imaging geometry invariant to object position and tilt, thereby improving pose determination accuracy without increasing processing complexity

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

Instead of correcting perspective distortion through computational methods after image capture, the patent inverts the approach by using optical design (telecentric lenses) to prevent perspective distortion from occurring in the first place. This shifts the solution from the digital processing domain to the optical domain

Inventive Principle:

Principle #13The other way round (Inversion)

2Length of stationary object

If small baseline distances are used between cameras, then the system compactness increases, but the accuracy of correspondence finding decreases requiring more textured features

Engineering Contradiction:

Improvecamera baseline distanceVSAvoidcorrespondence finding accuracy

Core Design Contradiction:

Length of stationary objectVSMeasurement precision

Solution Approach 1:

The patent changes the imaging geometry parameter by using telecentric lenses which provide parallel projection instead of perspective projection. This allows the system to maintain accurate correspondence finding even with small camera baselines, as the telecentric geometry eliminates the need for complex disparity calculations and reduces sensitivity to baseline variations

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If conventional perspective cameras are used, then the system can acquire images of objects at various distances, but the feature contrast and shape consistency deteriorate as viewing angle changes

Engineering Contradiction:

Improveviewing angle rangeVSAvoidfeature contrast consistency

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent inverts the conventional imaging approach by using telecentric lenses that provide view-invariant imaging. Instead of accepting perspective distortion as inevitable and trying to correct it computationally, the optical design itself ensures that features maintain consistent contrast and shape across different viewing angles and object positions

Inventive Principle:

Principle #13The other way round (Inversion)

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 enhances the speed and accuracy of pose determination, allowing for more precise guidance of manipulators in microscopic and near-microscopic applications, thereby improving throughput and efficiency in robotic operations.

Implementation Method 1

The use of non-perspective lenses, such as telecentric lenses, in conjunction with multiple cameras allows for the acquisition of images that are less sensitive to object tilt and position changes

Methodology Applied

Scientific EffectTelecentric lens optical principle: Lens

Data Source

PatentUS9124873B2System and method for finding correspondence between cameras in a three-dimensional vision system

Publication Date: 2015.09.01 COGNEX CORP

AI summary

This invention provides a system and method for determining correspondence between camera assemblies in a 3D vision system implementation having a plurality of cameras arranged at different orientations with respect to a scene involving microscopic and near microscopic objects under manufacture moved by a manipulator, so as to acquire contemporaneous images of a runtime object and determine the pose of the object for the purpose of guiding manipulator motion. At least one of the camera assemblies includes a non-perspective lens. The searched 2D object features of the acquired non-perspective image, corresponding to trained object features in the non-perspective camera assembly can be combined with the searched 2D object features in images of other camera assemblies, based on their trained object features to generate a set of 3D features and thereby determine a 3D pose of the object.