Automated Chart Data Extraction via Local Screenshot Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for extracting data from digital chart images are often inefficient, leading to poor data quality, erroneous data extraction, and require significant user intervention due to the complexity and diversity of chart types, especially on small display devices.
Innovation Solution
A method and system that uses a graphical user interface to enable users to select a screen area containing a chart, automatically takes a screenshot, performs image analysis, and outputs the extracted data with minimal manual steps, allowing for the extraction of numerical data from charts using OCR techniques and reducing man-machine interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated data extraction from digital chart images is performed using existing methods, then data extraction speed is improved, but data quality deteriorates due to poor extraction accuracy and erroneous data
Solution Approach 1:
The patent segments the chart image analysis into distinct phases: preprocessing the image, detecting chart elements (axes, bars, pie segments), extracting text labels, and correlating visual elements with textual data. This segmentation allows each phase to be optimized independently, improving both speed and accuracy of the overall extraction process
Solution Approach 2:
The patent introduces an intermediary processing layer that captures the display screen containing the chart, performs image analysis locally, and then outputs structured data. This intermediary approach avoids direct file processing while maintaining accuracy, and enables automated extraction without requiring manual intervention
2Loss of time
If existing automated extraction methods are used, then extraction time is reduced, but user intervention requirements increase due to need for manual correction of errors
Solution Approach 1:
The patent implements feedback mechanisms where the system analyzes extracted data for consistency and completeness, automatically correcting obvious errors. The system provides feedback loops that verify chart element detection accuracy and validate data extraction, reducing the need for manual correction while maintaining high automation levels
Solution Approach 2:
The patent performs preliminary actions by pre-processing the chart image (enhancement, normalization, feature detection) before main extraction. It also pre-identifies chart type and structure, which prepares the system for optimized extraction and reduces subsequent manual intervention needs
3Extent of automation
If existing extraction programs are used, then automation level is improved, but adaptability deteriorates as they work only on specific chart types
Solution Approach 1:
The patent creates a universal extraction system that can handle multiple chart types (bar charts, pie charts, line charts, mixed charts) through a single integrated platform. The system uses generic image analysis techniques that adapt to different chart structures, enabling one tool to serve multiple chart formats without requiring type-specific programs
4Power
If remote image analysis servers are used, then processing capability is improved, but security deteriorates due to upload of confidential data
Solution Approach 1:
The patent extracts the image analysis functionality from remote servers and implements it locally on the user's device. By taking out the processing capability from the remote environment and embedding it locally, the system maintains strong processing power while eliminating the security risk of uploading confidential chart data to external servers
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly reduces manual efforts and errors in data extraction, enabling efficient data capture from various chart types without the need for uploading images to remote servers, and allows for accurate data export to target applications with minimal user interaction.
Implementation Method 1
automatically performing, by the capture module, an image analysis of the screenshot for extracting data from the chart contained in the screenshot
Data Source
AI summary
A method for automatically extracting data from a chart includes generating and displaying a graphical user interface (GUI) enabling a user to select a screen area covering one or more screens at least partially, the screen area displaying a chart; in response to determining that the user has selected the screen area, automatically taking a screenshot of the selected screen area and automatically performing an image analysis of the screenshot for extracting data from the chart contained in the screenshot; in response to determining that the image analysis of the screenshot has completed successfully, automatically displaying or enabling a selectable GUI element; and in response to a select action of the user having selected said selectable GUI element, outputting the extracted data.