Improved local information-based CV model image segmentation method
A local information and image segmentation technology, which is applied in image analysis, image data processing, instruments, etc., can solve the problems of the model falling into local minimum, large amount of calculation, and uneven grayscale images, so as to improve the convergence speed and high Segmentation accuracy, the effect of enhancing robustness
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Example Embodiment
[0031] Example 1
[0032] The model segmentation speed based on global information is fast, but it cannot successfully segment the gray-level uneven image. Models based on local information tend to fall into local minima, leading to over-segmentation. The hybrid model that combines global and local information can segment grayscale heterogeneous images better, but the calculation is more complicated. In view of this situation, the present invention conducts research and proposes an image segmentation method based on improved local information CV model, see figure 1 , Including the following steps:
[0033] (1) Input the original image I, calculate the pixel gray value I(x), and x is the image pixel. The original image I is the image to be divided.
[0034] (2) Set the initial contour C of the original image I. The initial contour C is the image pixel point x corresponding to the arbitrarily designated closed curve of the image, which is obtained by the zero level set of the level ...
Example Embodiment
[0046] Example 2
[0047] The image segmentation method based on the CV model of improved local information is the same as that of embodiment 1, the global target gray scale fitting value c of the original image I described in step (6a) 1 And global background grayscale fitting value c 2 , The local target gray-scale fitting value f of the original image I 1 (x) and local background gray scale fitting value f 2 (x), calculated as follows:
[0048]
[0049] Among them, the Hyde function Generalized Gaussian function Γ(·) is the Gamma function, α is the scale parameter, β is the shape parameter, and * is the convolution operation.
[0050] The present invention introduces local information into the CV model based on global information, where the Gaussian function is extended to a generalized Gaussian function, and the scale parameter α and shape parameter β are introduced to make the image smoother and have better robustness against noise. Successfully segment grayscale heterogeneous...
Example Embodiment
[0051] Example 3
[0052] The image segmentation method of the CV model based on the improved local information is the same as that of the embodiment 1-2, the weighted target gray scale fitting value m of the original image I in step (6b) 1 (x) and weighted background gray fitting value m 2 (x), calculated by the following formula:
[0053] m i (x)=w·c i +(1-w)·f i (x), i=1, 2
[0054] Where w is the global gray-scale fitting coefficient, w∈[0,1], when the image gray is uniform, w takes the larger value, when the image gray is uneven, w takes the smaller value, and the specific values are combined The image is ok.
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap