Robot Picking of Deformed Objects Using Adaptive Pattern Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robot systems struggle to accurately pick up target objects that deform due to their placement state, such as bag-packaged bodies, as the deformation causes a mismatch in the recognized pattern, leading to potential failure in detection and pickup.
Innovation Solution
A robot system equipped with an imaging apparatus for capturing images, an image processing apparatus with a teaching unit, recognition unit, and correction unit to set and adjust a search range for feature extraction, allowing for accurate detection and pickup of deformed target objects by adjusting the search range based on the deformation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the robot system uses a fixed pattern recognition method based on uncompressed object shape, then it can accurately detect objects in their original state, but it fails to detect deformed objects that have been stacked and compressed
Solution Approach 1:
The search range is dynamically adjusted based on the deformation state of the target object. The correction unit modifies the search range parameters (such as width and height) according to the detected deformation degree, allowing the recognition system to adapt to different placement states and accurately detect both compressed and uncompressed objects
Solution Approach 2:
The system changes the parameters of the search range (width, height, position) based on the deformation characteristics of the target object. By adjusting these parameters according to the actual deformation state, the system maintains accurate pattern recognition despite variations in object shape due to stacking or compression
Data Source
AI summary
Provided is a robot system capable of appropriately removing an object that may deform depending on a placement state. A robot system according to the present disclosure is provided with: an imaging apparatus that captures an image of an object; a robot that picks up the object; an image processing apparatus that identifies the position of the object on the basis of the image captured by the imaging apparatus; and a robot controller that causes the robot to pick up the object, the position of which has been identified by the image processing apparatus. The image processing apparatus comprises: a teaching unit that sets a search range in which a feature amount for a pattern on the object in the image captured by the imaging apparatus may be obtained; a recognition unit that extracts the pattern having the feature amount within the search range in the image captured by the imaging apparatus; and a correction unit that corrects the search range on the basis of the feature amount of the pattern extracted by the recognition unit.