Robotic Gripper Vision System for Unstructured Object Handling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Designing a gripper that can reliably grip objects of varying shapes and sizes from an unstructured environment, especially when they are among other objects or obstructions, is challenging due to the need for prior knowledge of object shapes and orientations.

Innovation Solution

A method involving image processing to calculate masks for background, object, gripper, and gripping areas, determining overlapping areas, and selecting a gripping location based on quality factors to ensure effective gripping, even for objects larger than the gripper opening with protruding handles.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a generic gripper design is used to handle objects of varying shapes and sizes, then the gripper can be applied to unstructured environments, but the gripping reliability decreases

Engineering Contradiction:
Improvegripper adaptabilityVSAvoidgripping reliability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The system performs preliminary actions by calculating background masks, object masks, gripper masks, and gripping area masks before selecting a gripping location. These preliminary calculations enable the generic gripper to adapt to various objects while maintaining reliability through systematic pre-analysis of the gripping scenario.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The invention changes parameters by calculating quality factors based on different mask overlaps and selecting gripping locations that optimize these parameters. This allows the generic gripper to adapt to varying object shapes and sizes while maintaining reliable gripping through quantitative parameter optimization.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If the robot system uses complex image processing and mask calculations to determine gripping locations, then gripping accuracy improves, but processing time increases

Engineering Contradiction:
Improvegripping location precisionVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The image processing is segmented into distinct mask calculations (background mask, object mask, gripper mask, gripping area mask), allowing parallel processing and reducing overall computation time while maintaining high precision in gripping location determination.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs partial actions by calculating only the necessary mask overlaps relevant to gripping locations rather than analyzing entire images in detail, reducing processing time while maintaining sufficient precision for gripping decisions.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2658691B1Method, computer program and apparatus for determining a gripping location
Publication Date: 2017.09.13 ZENROBOTICS OY
  • EP2658691B1 patent drawingFigure 1
  • EP2658691B1 patent drawingFigure 2
  • EP2658691B1 patent drawingFigure 3

AI summary

According to one aspect of the invention, there is provided a method com- prising: obtaining at least one image comprising at least one object; analysing the at least one image to determine at least one gripping location to grip an object; selecting a gripping location from the at least one gripping lo- cation based on a predetermined criterion; and issuing at least one instruction to a gripper to grip the object at the selected gripping location.