Image-Based Grasp Control for Object State-Dependent Picking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot grasping systems fail to execute grasping operations based on the specific state of an object, leading to inefficient and potentially unsuccessful grasping due to unconditional reliance on input images and three-dimensional data.
Innovation Solution
An information processing apparatus that captures images, recognizes the object's state, and generates specific grasping operations by comparing the recognized state with predefined conditions, allowing the grasping unit to execute the optimal operation based on the detected state and conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If unconditional grasping is performed based on input images and three-dimensional data, then the grasping operation can be executed without complex state recognition, but the grasping cannot be adapted to different object states leading to reduced success rate
Solution Approach 1:
The system performs preliminary recognition of the object state (such as orientation, position, and shape) before executing the grasping operation. This preliminary action allows the system to adapt the grasping parameters based on the recognized state, thereby improving the grasping success rate without requiring complex real-time adjustments during the grasping process itself.
Solution Approach 2:
The system dynamically adjusts grasping parameters based on the recognized object state. Instead of using fixed grasping parameters, the system modifies the grasping operation according to the detected orientation, position, and shape of the object, enabling adaptive grasping that improves reliability while maintaining manageable system complexity through parameter adjustment rather than structural complexity.
2Adaptability or versatility
If state recognition is added to adapt grasping operations to object conditions, then the grasping can be executed in accordance with object state, but the system complexity increases
Solution Approach 1:
The system segments the state recognition process into distinct functional components that analyze specific aspects of the object (orientation, position, shape) separately. This segmentation allows the system to achieve comprehensive adaptability by combining results from multiple simple recognition modules rather than using a single complex recognition system, thereby improving adaptability while controlling system complexity.
Solution Approach 2:
The state recognition system is designed with multi-functional capabilities that can identify various object attributes (orientation, position, shape) using a unified recognition framework. This universal approach allows the same recognition system to handle different object states and types, improving grasping adaptability across diverse scenarios while avoiding the need for separate specialized systems for each function.
Data Source
AI summary
It is an object to enable a grasping operation to be executed according to a state of an object. The invention provides an information processing apparatus which determines the grasping operation in a grasping unit for grasping the object. The information processing apparatus has: an obtaining unit for obtaining an image acquired by capturing the object; a recognizing unit for recognizing a state of the object from the image obtained by the obtaining unit; and a generating unit for generating information for allowing the grasping unit to execute the grasping operation on the basis of the object state recognized by the recognizing unit and conditions to execute the grasping operation.