Camera-Based Object Segmentation for Unknown Workpiece Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for object instance recognition in dynamic work environments, such as palletizing and depalletizing, are inefficient and require manual intervention, particularly when dealing with a priori unknown objects.
Innovation Solution
A method and apparatus that utilize a camera system to detect a predefined starting region within an object instance, automatically segmenting it using a monitoring and control unit, without manual interaction, and determining geometric features for further processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual marking of positive and negative points is used to provide advance information for segmentation, then the quality of segmentation is improved, but the process becomes time-consuming and inefficient
Solution Approach 1:
The system automatically detects starting regions and performs segmentation without requiring manual marking by specialists. The monitoring and control unit autonomously identifies objects and segments them, making the system self-sufficient and eliminating the time-consuming manual intervention step while maintaining segmentation quality.
Solution Approach 2:
The system performs preliminary automatic detection of starting regions before segmentation. By pre-identifying potential starting regions through automated analysis of the captured image, the system prepares the segmentation process in advance without manual input, thus improving both speed and efficiency while maintaining accuracy.
2Measurement precision
If stationary 2D and/or 3D sensor systems are used to recognize goods on belts and pallets, then object recognition is achieved, but the system cannot efficiently handle a priori unknown objects without manual intervention
Solution Approach 1:
The monitoring and control unit autonomously detects starting regions and segments objects without requiring pre-programmed knowledge or manual configuration. This self-service capability enables the system to adapt to any object type automatically, significantly improving versatility for handling a priori unknown objects while maintaining recognition accuracy.
Solution Approach 2:
The system dynamically adapts to different object types by automatically detecting starting regions based on the actual image content rather than relying on fixed, pre-defined parameters. This dynamic approach allows the system to flexibly handle various unknown objects while maintaining precise recognition and segmentation.
3Manufacturing precision
If algorithms based on edge detection or CAD matching are used for segmentation, then segmentation can be performed, but the process is time-consuming and not efficient for automated tasks
Solution Approach 1:
The system performs automated detection of starting regions and segmentation without requiring complex edge detection algorithms or CAD matching processes. The monitoring and control unit autonomously identifies objects and segments them using simplified automated logic, thereby maintaining segmentation capability while dramatically improving automation efficiency and productivity.
Solution Approach 2:
The system extracts only the essential starting region information needed for segmentation, avoiding the computationally intensive processes of full edge detection or CAD matching. By taking out only the critical detection step, the system achieves efficient automated segmentation with improved productivity while preserving necessary segmentation precision.
Data Source
AI summary
The invention relates to a method for recognizing at least one instance of an object during a work sequence in a working environment, said method comprising the steps (A) recording an image of the working environment by means of a camera apparatus, (B) transmitting the image to a monitoring and control unit, (C) detecting a predefined or predefinable starting region for the segmenting of the instance in the image, whereby the instance is selected by the monitoring and control unit for the segmenting, and (D) segmenting the instance in the image, starting from the starting region that is detected and that is arranged within the instance, by means of the monitoring and control unit and recognizing the segmented instance.