Image Segmentation via Super-Pixel Distance Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image segmentation methods suffer from under-segmentation and over-segmentation, particularly in images with adhered targets, leading to lower accuracy and efficiency due to limitations in binary image processing.
Innovation Solution
The method performs super-pixel segmentation, binarization, distance transformation, and seed point marking to obtain a segmented image, effectively distinguishing target regions and boundary pixel blocks by utilizing a ratio of pixel values and distance transformation to create a grayscale super-pixel image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If binary image processing is used for image segmentation, then processing speed is improved, but segmentation accuracy deteriorates due to under-segmentation and over-segmentation
Solution Approach 1:
The patent applies super-pixel segmentation to divide the image into larger super-pixels before binary processing, which reduces the number of pixels that need to be processed while maintaining segmentation accuracy. This prevents both under-segmentation and over-segmentation by creating an intermediate representation that balances detail preservation with processing efficiency
Solution Approach 2:
The patent introduces a grayscale dimension to the binary super-pixel image through distance transformation, creating a grayscale super-pixel image that contains information about pixel block positions and relationships. This additional dimension enables accurate contour curve location while maintaining the simplified super-pixel structure for efficient processing
2Productivity
If super-pixel segmentation is performed, then processing efficiency is improved, but boundary precision deteriorates
Solution Approach 1:
The patent performs preliminary super-pixel segmentation and binarization before the main segmentation process. This preliminary action creates a simplified representation that guides subsequent processing, allowing for efficient computation while maintaining boundary precision through the distance transformation step that preserves spatial relationships
Solution Approach 2:
The patent introduces a grayscale super-pixel image as an intermediary representation between the binary super-pixel image and the final segmented image. This intermediary contains enhanced boundary information through distance transformation, enabling precise boundary detection while maintaining the computational efficiency of super-pixel processing
Data Source
AI summary
An image segmentation method and device, a computer device and a non-volatile storage medium are provided. The image segmentation method includes: performing super-pixel segmentation on an image to be segmented to obtain a super-pixel image, and binarizing the image to be segmented to obtain a binary image; combining the super-pixel image and the binary image to obtain a binary super-pixel image; performing distance transformation on the binary super-pixel image, to obtain a grayscale super-pixel image; marking seed points in the grayscale super-pixel image, to obtain a seed point super-pixel image in which grayscale values of the seed points are greater than a first value, and grayscale values of pixel blocks other than the seed points in target regions are the first value; and marking and filling the pixel blocks, grayscale values of which are the first value in the seed point super-pixel image, to obtain a segmented image.