Shelf Depth Imaging for Accurate Robot Picking and Placing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The difficulty of picking or placing items by transport robots is increased due to items being placed on shelves, and factors such as shelf vibration and manual operation errors compromise safety and efficiency.
Innovation Solution
A photographing method and system for transport robots that utilize multi-dimensional image information to determine the presence and identity of items on shelves, allowing for precise picking or placing operations, including strategies for handling obstacles and errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple items are placed on a shelf to increase storage capacity, then the quantity of stored items is improved, but the difficulty of picking or placing items increases
Solution Approach 1:
The patent introduces depth direction photographing capability to the shelf system, transforming a 2D shelf surface into a 3D searchable space. The photographing module captures images along the depth direction, enabling the robot to locate items at different depths through multi-dimensional image processing, thus solving the difficulty of picking/placing items in densely packed shelves
2Reliability
If shelf vibration and manual operation errors occur in real scenarios, then the robustness of the system is tested, but the safety of picking or placing cannot be ensured
Solution Approach 1:
The patent implements a feedback mechanism where the photographing module continuously captures images of the shelf and target items. The image processing unit analyzes these images to determine item positions and states, providing real-time feedback to the control unit to adjust picking/placing operations, thereby ensuring safety despite vibrations or manual errors
Solution Approach 2:
The patent replaces manual operation with an automated robot system equipped with photographing and image processing capabilities. This substitution eliminates manual errors and provides precise, repeatable operations that can compensate for shelf vibrations through real-time image-based positioning
3Measurement precision
If multi-dimensional image information is obtained along the depth direction of the shelf, then the accuracy of item identification is improved, but the complexity of the photographing system increases
Solution Approach 1:
The patent segments the photographing task into multiple directional components (depth direction, width direction, height direction). The photographing module captures images along the depth direction specifically, and the image processing unit processes these segmented images to extract item position information, making the complex 3D positioning task manageable through systematic segmentation
Data Source
AI summary
A photographing method for picking or placing, applied includes: obtaining first multi-dimensional image information of a target position in a target shelf; determining, according to the first multi-dimensional image information, whether there is a first item in the target position; and determining a photographing strategy of the photographing module according to a determining result, wherein the photographing strategy includes one of: the photographing module not moving with the handling apparatus in a telescopic direction for continued photographing, the photographing module moving a preset distance along the telescopic direction with the handling apparatus, and performing an operation of starting a solution; the solution includes at least one of stopping photographing, sending a warning signal, and reporting to a server to which the transport robot belongs.