Adaptive Border Detection via User Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic page border detection algorithms are not accurate when the image background contains clutter, edges, or lacks sufficient contrast between the document and the background, leading to inaccurate border detection and increased manual intervention.
Innovation Solution
A method and system that dynamically learn from user corrections by adjusting default parameters for automatic border detection, allowing for improved border detection accuracy by presenting the detected border to the user, receiving adjustments, and updating parameters based on these corrections for future image captures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If automatic border detection algorithms are applied to images with cluttered backgrounds or low contrast, then the detection process is automated, but the detection accuracy deteriorates
Solution Approach 1:
The system displays the detected border to the user and receives feedback through manual adjustments. These adjustments are then used to update the default parameters for future automatic detections, creating a continuous improvement loop that enhances accuracy while maintaining automation.
Solution Approach 2:
The system dynamically updates default parameters based on user corrections. By changing parameters such as edge detection thresholds, contrast sensitivity, and adaptive lighting compensation settings based on feedback from manual adjustments, the system adapts to different image conditions and improves detection accuracy.
2Productivity
If default parameters are used for automatic border detection, then the process is fast and simple, but the detection accuracy deteriorates in challenging image conditions
Solution Approach 1:
The system performs preliminary automatic border detection using default parameters to quickly identify potential borders. This preliminary action provides a starting point that can be rapidly refined based on user feedback, maintaining speed while improving accuracy.
Solution Approach 2:
The system automatically updates its own default parameters based on user corrections without requiring manual reconfiguration. This self-service mechanism allows the system to learn from user interactions and improve its performance automatically while maintaining operational speed.
3Measurement precision
If manual adjustment of borders is allowed, then detection accuracy can be improved, but the ease of operation deteriorates due to additional user intervention
Solution Approach 1:
The system presents the automatic detection result to the user and accepts feedback in the form of manual adjustments. This feedback mechanism allows users to correct inaccuracies without requiring complete manual redrawing, maintaining ease of operation while improving accuracy.
Solution Approach 2:
The system automatically applies parameter updates based on user adjustments without requiring users to manually configure complex parameters. This self-service approach simplifies the user experience while still incorporating manual correction capabilities for improved accuracy.
4Measurement precision
If user corrections are used to update default parameters, then future detection accuracy improves, but the device complexity increases due to parameter management
Solution Approach 1:
The system automatically manages parameter updates based on user corrections without requiring manual configuration or complex parameter management interfaces. This self-service approach handles the complexity internally while maintaining simplicity for the user.
Solution Approach 2:
The system dynamically adjusts default parameters based on aggregated user corrections from multiple images. By automatically managing parameter changes and learning from patterns in user adjustments, the system improves future detection accuracy without exposing users to parameter management complexity.
Data Source
AI summary
A method of automatically identifying a border in a captured image may include capturing an image of a target by an image sensor of an electronic device, and, by one or more processors, processing the image to automatically detect a border of the target in the image by applying an automatic border detection method to the image. The method may include presenting the image of the target to a user via a display of the electronic device so that the presented image comprises a visual depiction of the detected border, receiving an adjustment of the border from the user, determining whether to update the default parameters based on the received adjustment, in response to determining to update the default parameters, determining one or more updated parameters for the automatic border detection method that are based on, at least in part, the received adjustment, and saving the updated parameters.