Character Detection and Binarization via Segmented Resolution Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems require large memory capacity and extensive calculations to achieve high-quality document computerization, which is impractical in limited resource environments, as data compression compromises image quality.
Innovation Solution
An image processing apparatus with a resolution converting unit, object dividing unit, and decompressing unit that generates high-resolution character contours by processing compressed images with low-resolution character objects, allowing for efficient vector image creation in small memory capacity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data compression is applied to reduce memory capacity requirements, then memory usage is reduced, but image quality deteriorates
Solution Approach 1:
The image processing is segmented into two resolution levels: low-resolution processing for character object detection and high-resolution processing for contour generation. This allows the system to work with compressed low-resolution data for object identification while using uncompressed high-resolution data for quality contour extraction, thus resolving the contradiction between memory usage and image quality.
Solution Approach 2:
Character objects are detected and identified in advance using low-resolution compressed images. This preliminary action allows the system to know which regions require high-resolution processing, enabling selective decompression and contour generation only for relevant character regions, thereby reducing overall memory requirements while maintaining quality where needed.
2Manufacturing precision
If high-resolution processing is applied to maintain image quality, then image quality is improved, but processing time increases
Solution Approach 1:
The processing pipeline is divided into fast low-resolution character detection and slower high-resolution contour generation. By segmenting the workload, the system performs quick identification at low resolution and only applies computationally intensive high-resolution processing to identified character regions, thus improving overall processing speed while maintaining quality output.
Solution Approach 2:
The system applies high-resolution processing only partially - specifically to character object regions identified in the low-resolution image - rather than processing the entire image at high resolution. This partial action maintains image quality for character contours while significantly reducing processing time compared to full high-resolution processing.
3Manufacturing precision
If extensive calculations are performed to achieve high-quality vectorization, then vectorization quality is improved, but processing time increases
Solution Approach 1:
Character objects are preliminarily identified using simple low-resolution image analysis. This preliminary step provides accurate character region information that guides subsequent high-quality vectorization, avoiding the need for extensive calculations on the entire image and reducing processing time while maintaining vectorization quality.
Solution Approach 2:
Complex vectorization calculations are performed only on character object regions identified in advance, rather than on the entire image. This partial application of intensive processing maintains high vectorization quality for text elements while significantly reducing overall processing time.
Data Source
AI summary
There is provided an image processing apparatus which can generate a vector image with a high image quality at a high speed in a small memory capacity. The present invention generates an image with low resolution by performing resolution conversion to a compressed image with high resolution and obtains information of a character object with low resolution by performing object division to the image with the low resolution. The compressed image with the high resolution is decompressed and an image showing a character contour with high resolution is generated by using the information of the character object with the low resolution obtained by the object dividing and the decompressed image with the high resolution. At the time of performing the decompression, the compressed image with the high resolution may be partially decompressed based upon the information of the character object.