Image Binarization via Slicing and Stitching for Edge Preservation

Solution Overview

Problem

In digital image processing, existing methods face challenges in balancing processing speed and image quality, particularly when dealing with large original images, as image compression ratios often compromise either speed or quality during binarization processing.

Innovation Solution

The method involves slicing the original image into multiple segments, processing each slice using a binarization model, stitching the results to form a binary image, and then synthesizing it with another binary image to enhance contour detection, thereby improving processing speed and image quality without compressing the original image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If image compression is applied during binarization processing, then processing speed is improved, but image quality and edge details deteriorate

Engineering Contradiction:

Improveprocessing speedVSAvoidimage quality

Core Design Contradiction:

ProductivityVSManufacturing precision

Solution Approach 1:

The original image is divided into multiple sub-images through slicing processing, allowing parallel or sequential processing of smaller segments. This segmentation enables faster processing while maintaining the ability to preserve edge details in each segment through the binarization model, thus resolving the contradiction between processing speed and image quality.

Inventive Principle:

Principle #1Segmentation

2Manufacturing precision

If the original image is processed without slicing, then image quality is maintained, but processing time increases

Engineering Contradiction:

Improveimage qualityVSAvoidprocessing time

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The original image is divided into multiple sub-images through slicing processing, allowing parallel or sequential processing of smaller segments. This segmentation enables faster processing while maintaining the ability to preserve edge details in each segment through the binarization model, thus resolving the contradiction between processing speed and image quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

After processing individual sub-images through binarization, the results are merged through stitching processing to reconstruct the complete binarized image. This merging step ensures that the final output maintains the quality and detail of the original image while benefiting from the speed improvements of processed segments.

Inventive Principle:

Principle #5Merging (Combining)

3Quantity of substance

If image binarization is performed on large images, then data reduction is achieved, but processing complexity increases

Engineering Contradiction:

Improvedata amountVSAvoidprocessing complexity

Core Design Contradiction:

Quantity of substanceVSDevice complexity

Solution Approach 1:

The original image is divided into multiple sub-images through slicing processing, allowing parallel or sequential processing of smaller segments. This segmentation enables faster processing while maintaining the ability to preserve edge details in each segment through the binarization model, thus resolving the contradiction between processing speed and image quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

After processing individual sub-images through binarization, the results are merged through stitching processing to reconstruct the complete binarized image. This merging step ensures that the final output maintains the quality and detail of the original image while benefiting from the speed improvements of processed segments.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS11985287B2Image processing method, image processing device, electronic apparatus and storage medium

Publication Date: 2024.05.14 HANGZHOU GLORITY SOFTWARE LTD

AI summary

An image processing method, an image processing device, an electronic apparatus, and a non-transitory computer-readable storage medium are provided. The image processing method includes: obtaining an original image; performing slicing processing on the original image to obtain multiple slice images of the original image; respectively processing the slice images through a first binarization model to obtain multiple slice binary images respectively corresponding to the slice images; performing stitching processing on the slice binary images to obtain a first binary image; processing the first binary image to obtain a pixel circumscribed contour image; processing the original image through a second binarization model to obtain a second binary image; synthesizing the second binary image and the first binary image according to positions of multiple circumscribed contour pixels in the pixel circumscribed contour image to obtain a synthetic image.