Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Object-based OBIA-SVM-CNN remote sensing image classification method

A technology of OBIA-SVM-CNN and RBF-SVM, which is applied in the direction of instruments, character and pattern recognition, and computer parts, can solve the problems of low accuracy of remote sensing classification and recognition, and improve the accuracy of remote sensing classification, improve classification accuracy, The effect of overall classification accuracy improvement

Active Publication Date: 2019-05-10
NORTHEAST INST OF GEOGRAPHY & AGRIECOLOGY C A S
View PDF2 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the problem of low accuracy of remote sensing classification and recognition in existing complex farmland areas, and propose an object-based OBIA-SVM-CNN remote sensing image classification method

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Object-based OBIA-SVM-CNN remote sensing image classification method
  • Object-based OBIA-SVM-CNN remote sensing image classification method
  • Object-based OBIA-SVM-CNN remote sensing image classification method

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0022] Specific implementation mode one: the specific process of the object-based OBIA-SVM-CNN remote sensing image classification method in this implementation mode is as follows: figure 1 shown.

[0023] Step 1. Segment the remote sensing image based on the multi-scale segmentation algorithm until all the segmented objects match the border of the farmland patch on the remote sensing image by visual inspection;

[0024] Step 2, train the RBF-SVM model and the CNN model, and obtain the trained RBF-SVM model and the CNN model;

[0025] Step 3. Use the trained RBF-SVM model and CNN model to classify and predict each segmented object in step 1 (for example, whether the segmented object is a farmland patch or a non-farmland patch), and obtain the CNN model prediction result and the SVM model forecast result;

[0026] Step 4: Carry out decision-making fusion of the prediction results of the CNN model and the prediction results of the SVM model to obtain the final classification r...

specific Embodiment approach 2

[0027] Specific embodiment 2: The difference between this embodiment and specific embodiment 1 is that in the first step, the remote sensing image is segmented based on the multi-scale segmentation algorithm, until all the segmented objects and farmland patches on the remote sensing image are visually inspected Boundary matching; the specific process is:

[0028] The segmented object is a farmland patch;

[0029] With the support of Yikang software, a multi-resolution segmentation algorithm (multi-resolution segmentation, MRS) is used to segment remote sensing images into farmland patches (objects) with homogeneous spectral and spatial information;

[0030] The multi-scale segmentation algorithm contains three segmentation control parameters, namely scale (scale), color measure (color) and smoothness measure (smoothness);

[0031] Determine the optimal combination of segmentation control parameters by trial and error (manually adjust the scale, color scale, and smoothness sca...

specific Embodiment approach 3

[0033] Specific embodiment three: the difference between this embodiment and specific embodiment one or two is that in the step two, the RBF-SVM model and the CNN model are trained to obtain the trained RBF-SVM model and the CNN model; the specific process is :

[0034] Select the radial basis function (radial basis function, RBF) as the kernel function (RBF-SVM) of the support vector machine to establish the support vector machine model;

[0035] Select M training sample points with classification labels, and extract image features according to the training sample points (for example, the remote sensing images with classification labels are remote sensing images of farmland patches and non-farmland patches in known remote sensing images, or the classification labels are Dadu and corn) as the input of RBF-SVM model;

[0036] The penalty coefficient C and the kernel coefficient γ are used as parameters of the RBF-SVM model;

[0037] The training sample point is a pixel in the...

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

No PUM Login to View More

Abstract

The invention provides an object-based OBIA-SVM-CNN remote sensing image classification method, and relates to a remote sensing image classification method. The objective of the invention is to solvethe problem of low remote sensing classification and identification accuracy of an existing complex farmland area. The method comprises the following steps: 1, segmenting a remote sensing image basedon a multi-scale segmentation algorithm until all segmented objects are matched with farmland plaque boundaries on the remote sensing image through visual inspection; 2, trainning RBF-SVM model and CNN model to get trained RBF-SVM model and CNN model; 3, using the trained RBF-SVM model and the CNN model to classify and predict each segmented object in the first, and obtaining the prediction results of the CNN model and the prediction results of the SVM model.; and 4, performing decision fusion on the CNN model prediction result and the SVM model prediction result to obtain a final classification result. The method is applied to the field of remote sensing image classification.

Description

technical field [0001] The invention relates to a remote sensing image classification method. Background technique [0002] The real-time monitoring of the spatio-temporal distribution information of farmland is of great significance for the estimation of farmland yield and the protection of food security on a national scale and even on a global scale. Remote sensing technology has developed into one of the mainstream means of farmland monitoring and classification because it has many unique advantages, including macroscopicity, current situation, repeatability and economy. With the rapid development of modern remote sensing technology, users can now obtain a large number of high-resolution (high spatial resolution, HSR) remote sensing images, which provides new opportunities for many remote sensing applications (including farmland monitoring and classification). However, although high-resolution images have richer structural information and texture information, compared wi...

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
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62
Inventor 李华朋张策张淑清丁小辉吴鹏田亚君
Owner NORTHEAST INST OF GEOGRAPHY & AGRIECOLOGY C A S
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com