Workflow Object Segmentation Using Auto-Detected Output Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for object detection in dynamic workflows, such as palletizing and depalletizing, are inefficient and require manual interaction for segmentation, especially when dealing with a priori unknown objects.
Innovation Solution
A method and device that utilize a camera and control unit to automatically detect a source region within an object's image, using markers like QR codes, to segment and recognize instances without manual input, enabling efficient and cost-effective detection and geometric feature determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual marking of positive and negative points is used for segmentation, then segmentation quality is improved, but operation complexity and time consumption increase significantly
Solution Approach 1:
The system automatically detects instances and performs segmentation without requiring manual marking of positive and negative points. The control unit processes images autonomously to identify and segment instances, eliminating the need for specialist intervention and making the system self-sufficient in the segmentation task.
Solution Approach 2:
The manual mechanical process of marking points by specialists is replaced with an automated image processing system using a camera device and control unit. The system substitutes human operation with automated detection algorithms that analyze images to identify instances and perform segmentation automatically.
2Ease of operation
If automated instance segmentation is implemented without preliminary information, then ease of operation is improved, but segmentation precision deteriorates
Solution Approach 1:
The system performs preliminary detection of instances in the image before segmentation. The control unit first identifies potential instances by detecting specific patterns or features in the captured image, then uses this preliminary information to guide the subsequent segmentation process, ensuring both automation and precision.
Solution Approach 2:
The system introduces an intermediary detection stage between image capture and final segmentation. This intermediary step automatically identifies instances and extracts relevant features, serving as a bridge that enables automated operation while maintaining segmentation precision through structured intermediate results.
3Adaptability or versatility
If traditional segmentation methods are used for unknown objects, then adaptability is reduced, but device complexity is increased due to manual intervention requirements
Solution Approach 1:
The system autonomously adapts to unknown objects by automatically detecting and analyzing their characteristics in the captured images. No manual configuration or marking is needed, allowing the system to handle diverse, previously unseen objects while maintaining simplicity in operation and reducing overall system complexity.
Solution Approach 2:
The system dynamically adjusts its detection and segmentation parameters based on the characteristics of each detected instance. Rather than requiring pre-programmed knowledge of specific objects, the system adapts its behavior in real-time to accommodate unknown objects, enhancing versatility without increasing device complexity.
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for detecting at least one instance of an object during a workflow in a work environment, comprising the steps (A) capturing an image of the work environment by a camera device, (B) transmitting the image to a control unit, (C) detecting a predetermined or predefinable output region for segmenting the instance in the image, whereby the instance is selected for segmentation by the control unit, and (D) segmenting the instance in the image starting from the detected output region located within the instance by the control unit and detecting the segmented instance.