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

High spectral abnormity detecting method based on dynamic weight deep self-coding

A dynamic weight and anomaly detection technology, applied in the field of hyperspectral anomaly detection, can solve the problems of local model pollution and low detection accuracy, and achieve the effect of solving pollution and improving detection accuracy.

Active Publication Date: 2018-03-23
HARBIN INST OF TECH
View PDF9 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to solve the problem that the abnormal target pollutes the local model in the existing hyperspectral anomaly detection method, resulting in low detection accuracy, and proposes a hyperspectral anomaly detection method based on dynamic weight depth self-encoding

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
  • High spectral abnormity detecting method based on dynamic weight deep self-coding
  • High spectral abnormity detecting method based on dynamic weight deep self-coding
  • High spectral abnormity detecting method based on dynamic weight deep self-coding

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0030] Specific implementation mode one: combine figure 1 To illustrate this embodiment, the specific process of a hyperspectral anomaly detection method based on dynamic weight depth self-encoding in this embodiment is as follows:

[0031] Step 1. Input the original hyperspectral image data into the DBN model, train the parameters of the DBN model, and obtain the optimized DBN model;

[0032] Step 2, input the tested image into the optimized DBN model, encode the tested image, and obtain the coded image of the tested image and the corresponding reconstruction error image;

[0033] The measured image is an original hyperspectral image;

[0034] Step 3: Input the coded image obtained in step 2 into the local pixel coding selection block, take a pixel in the measured image as the measured pixel, and obtain the local coded image of the measured pixel for the measured pixel; perform step 5 ;

[0035] Step 4. Input the reconstruction error image obtained in step 2 to the local ...

specific Embodiment approach 2

[0040] Specific embodiment two: the difference between this embodiment and specific embodiment one is: in the step one, the original hyperspectral image data is input into the DBN model, the DBN model parameters are trained, and the optimized DBN model is obtained; the specific process is:

[0041] Step 11, constructing the DBN model, pre-training the DBN model, and obtaining the preliminary estimated value of the DBN model parameters;

[0042] The specific process is:

[0043] The DBN model (DBN neural network) is composed of a multi-layer RBM model, and the single-layer RBM model is as follows: figure 2 shown. Including n visible layers, m hidden layers, and the weight coefficient w connecting visible layers and hidden layers;

[0044] The visible layer v is the input of the RBM, which is a column vector of n×1, and each component corresponds to each spectral band of the hyperspectral image;

[0045] The hidden layer h is the output of RBM, which is a column vector of m×...

specific Embodiment approach 3

[0094] Specific embodiment three: the difference between this embodiment and specific embodiment one or two is that: in the step two, the tested image is input into the optimized DBN model, the tested image is encoded, and the coded image of the tested image is obtained and the corresponding reconstruction error image; the specific process is:

[0095] DBN Inference Image Encoding

[0096] It mainly obtains the image coding after the image passes through the DBN network. The original hyperspectral image is input to the DBN inference coding module, which is a standard DBN model, and its parameters are generated during the training phase of the DBN model. The model structure is as image 3 shown

[0097] Input the tested image to the optimized DBN model. The number of neurons in the middle layer of the DBN model is generally lower than the number of nodes in the input and output layers. The entire DBN model has a symmetrical structure, and the output of the neurons in the mid...

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 a high spectral abnormity detecting method based on dynamic weight deep self-coding and relates to a high spectral abnormity detecting method. The invention aims to settle a problem of low detecting precision caused by partial model pollution by an abnormal model in an existing high spectral abnormity detecting method. The method comprises the steps of 1, obtaining an optimized DBN model; 2, obtaining a coding image and a reconstruction error image; 3, obtaining a local coding image, and performing step 5; 4, obtaining a local reconstruction error set, and performing step 6; 5, obtaining a local distance factor, and performing step 7; 6, obtaining all dynamic weights of the local distance, and performing step 7; and 7, obtaining an abnormity detecting operator value,setting a threshold, and when the abnormity detecting operator value is larger than or equal with the threshold, determining the detected pixel as an abnormal target, and otherwise, determining the tested pixel as a background pixel; taking a next pixel in a detected image as the detected pixel, and performing the steps 3-7 until all pixels in the detected image are determined. The high spectralabnormity detecting method is used for a high spectral abnormity detecting period.

Description

technical field [0001] The invention relates to a hyperspectral anomaly detection method. Background technique [0002] With the continuous development and progress of remote sensing imaging technology, hyperspectral images are playing an increasingly important role in precision agriculture, urban planning, military investigation and other fields. The research and application of hyperspectral remote sensing images has been the focus of relevant researchers for a long time. the key of. Compared with visible light or infrared remote sensing imaging technology, hyperspectral images can not only obtain the spatial distribution information of ground objects, but also collect spectral information of tens to hundreds of continuous narrow bands corresponding to ground objects at each pixel point. The characteristics of unity, so that the information of surface objects and materials can be distinguished through spectral information. However, in practical applications, due to the la...

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): G06T7/00G06T7/194
CPCG06T7/00G06T2207/10032G06T2207/20081G06T7/194
Inventor 彭宇马宁王少军刘大同
Owner HARBIN INST OF TECH
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