Image Rectification via Edge Detection and Cropping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image readers face challenges in efficiently and automatically discriminating between different types of data, such as 1D and 2D barcodes, text, and logos, and in minimizing image file size while maintaining data integrity, especially when dealing with varying document sizes and orientations.
Innovation Solution
The system employs a method to automatically crop and rectify images by identifying regions of interest using nominally straight edges and energy mapping, binarizing images, and applying lossless compression, allowing for orientation correction and image resizing to focus on specific data regions, thereby reducing file size and improving data processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the entire image is captured and stored, then complete visual information is retained, but image file size increases
Solution Approach 1:
The patent extracts only the necessary portion of the image (the dataform region) from the entire captured image. The image processor identifies and isolates the dataform area, storing or transmitting only this extracted region rather than the complete image, thereby reducing file size while preserving essential information.
Solution Approach 2:
The patent segments the captured image into distinct regions, separating the dataform area from the rest of the document. This segmentation allows the system to process, store, or transmit only the relevant dataform portion, reducing overall data volume while maintaining data integrity.
2Measurement precision
If manual discrimination of data types is used, then accurate identification is achieved, but operation time increases
Solution Approach 1:
The patent implements automatic discrimination where the image processor autonomously identifies and categorizes different data types (barcodes, text, logos, etc.) without requiring manual intervention. The system self-services by automatically analyzing image features, patterns, and characteristics to determine data type, thereby eliminating manual discrimination time while maintaining accuracy.
Solution Approach 2:
The patent replaces manual mechanical discrimination with an automated image processing system. The mechanical action of manually examining and identifying data types is substituted by electronic image analysis algorithms that automatically detect and classify data forms based on their visual characteristics.
3Measurement precision
If high resolution imaging is used, then data reading accuracy is improved, but image file size increases
Solution Approach 1:
The patent extracts only the essential dataform region from the high-resolution captured image. By isolating and processing only this extracted region at high resolution while reducing or eliminating storage of the complete high-resolution image, the system maintains data reading accuracy while minimizing file size.
Data Source
AI summary
A method of operating an image reader typically includes: searching a digital image for nominally straight edges; characterizing the nominally straight edges in terms of length and/or direction; determining a predominant orientation of the nominally straight edges; establishing a group of edges as a function of their proximity to the center of the image; establishing a group of edges as a function of their proximity to other remaining edge positions; and transmuting a rectangle bounding those edges into a rectified image. The rectified image is typically an image that is cropped or rotated.