SAR Image Segmentation Method Based on Semantic Conditional Random Field Model
A conditional random field and image segmentation technology, which is applied in the field of image processing, can solve the problems of non-semantic consistency of segmentation results, loss of detailed information of segmentation results, affecting SAR image classification, recognition and detection, etc. The effect of regional consistency
Active Publication Date: 2019-05-03
XIDIAN UNIV
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, the binary potential function only captures the isotropic relationship in the image space context, ignoring the anisotropic relationship of the SAR image itself, resulting in the loss of detailed information in the segmentation results, and the segmentation results do not have semantic consistency. Affect the subsequent classification, identification and detection of SAR images
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0029] refer to figure 1 , the present invention is based on the region map of the SAR image, divides the SAR image into the subspace of the mixed aggregation structure, the subspace of the structure region and the subspace of the homogeneous region; feature, and then use the method of AP clustering to obtain the segmentation result of the feature subspace; construct a semantic conditional random field for the structural region subspace and homogeneous region subspace; the unary potential function of the semantic conditional random field adopts polynomial logic Sti regression function and the statistical characteristics of SAR images; the binary potential function of the semantic conditional random field is expressed by the polynomial logistic Sti regression function based on the mixed kernel function; The segmentation results of the quality region subspaces are merged to obtain the segmentation results of the SAR image. The specific implementation steps are as follows:
[00...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an SAR image segmentation method based on a semantic condition random field model, and mainly to solve the problem of failing to retain detailed information of an image in the prior art. The implementation steps include: 1. according to the region map of an SAR image, dividing the SAR image into a mixed aggregate structure surface feature subspace, a structural region subspace and a homogeneous region subspace; 2. extracting features for the mixed aggregate structure surface feature subspace using the bag-of-words model, and segmenting using an AP clustering method; 3. constructing a semantic condition random field model to segment the structural region subspace and the homogeneous region subspace; and 4. obtaining segmentation results of the SAR image by combining segmentation results of the mixed aggregate structure surface feature subspace, the structural region subspace and the homogeneous region subspace. The invention achieves good segmentation effect of the SAR image and can be used for semantic segmentation of the SAR image.
Description
technical field [0001] The invention belongs to the technical field of image processing, in particular to a SAR image segmentation method, which can be used for image classification, recognition and detection. Background technique [0002] Random field method is a popular method in SAR image segmentation. A typical random field method is the Markov random field MRF model, which is a probability generation model. In the MRF model, the posterior probability is equal to the product of the likelihood probability and the prior probability. Likelihood probability describes the characteristics of SAR images, and is usually represented by the statistical distribution of SAR images, and the choice of distribution is mainly based on the characteristics of SAR images. The prior probability describes the spatial context information of the image, usually represented by Gibbs distribution. However, strong dependencies are required in the assumptions of the MRF model, and the prior mode...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06T7/143
CPCG06T2207/10044G06T2207/20112
Inventor 刘芳段一平李婷婷焦李成郝红侠陈璞华马晶晶尚荣华
Owner XIDIAN UNIV