Unlock instant, AI-driven research and patent intelligence for your innovation.
Hyperspectral Feature Learning Method Based on Recursive Autoencoding
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A technology of automatic coding and feature learning, which is applied in the field of image processing, can solve problems such as difficult to find class labels, and achieve the effect of fast extraction of signs, economical and easy implementation, and high recognition accuracy
Active Publication Date: 2017-08-25
XIDIAN UNIV
View PDF1 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0006] The purpose of the invention is to overcome the problem that the class mark is relatively difficult to find in the existing hyperspectral feature extraction technology
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0044] In order to overcome the problem that class labels are difficult to find in the existing hyperspectral feature extraction technology, this embodiment provides a method such as figure 1 , image 3 and Figure 4 The shown hyperspectral feature learning method based on recursive auto-encoding includes the following steps:
[0045] (1) Input hyperspectral remote sensing image data, each pixel or sample is represented by a spectral feature vector, the feature dimension of the sample is d, and the sample set is normalized to be between 0 and 1;
[0046] (2) From the normalized sample set A certain proportion of the samples are selected as the training set, and the remaining samples are used as the test set, where x i is the i-th sample, N is the total number of samples, represents the field of real numbers;
[0047] (3) Construct the neighborhood window block of each sample: on the normalized sample set, take each sample as the center, take all the samples in its m×m n...
Embodiment 2
[0083] The effect of the present invention will be further described below in conjunction with FIG. 2(a) and FIG. 2(b).
[0084] The simulation experiment of this embodiment is implemented on an Intel Core(TM) 2 Duo CPU with a main frequency of 2.33GHz, a memory of 2G, and MATLAB 7.14 on a Windows 7 platform.
[0085] The simulation of this embodiment is carried out on two representative hyperspectral data Indian Pines and PaviaUniversity. The Indian Pines image contains 16 types of ground features: alfalfa, corn-unploughed, corn-irrigated, corn, pasture, trees, Cut pasture, haystack, buckwheat, soybean-unploughed, soybean-irrigated, soybean, wheat, forest, building-grass-tree, stone-rebar; Pavia University image contains 9 types of ground features: alfalfa, meadow , gravel, trees, colored flakes, bare soil, asphalt, bricks, shadows.
[0086] Figure 2(a) shows the Indian Pines hyperspectral data image, and Figure 2(b) shows the Pavia University hyperspectral data image. In g...
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 belongs to the technical field of image processing, and specifically provides a hyperspectral feature learning method based on recursive automatic coding. Training set and test set; construct a neighborhood window block for each feature vector; in the training set, each feature vector and its neighbors are merged according to the criterion of minimizing the reconstruction error between feature vectors, and then according to the entire network Reconstruct the error to train the entire network; respectively input the training set and test set into the trained network to obtain a new training set and test set; input the new training set and test machine to the support vector machine for classification, and obtain classification results. The invention adopts the idea of unsupervised, overcomes the problem of difficulty in acquiring hyperspectral data, obtains a higher classification accuracy rate, and can be used in the fields of mineral exploration, environmental management, and military defense.
Description
technical field [0001] The invention belongs to the technical field of image processing, in particular to a hyperspectral feature learning method based on recursive automatic coding in the technical field of remote sensing images. Background technique [0002] Hyperspectral image technology has developed rapidly in recent years, and its research is mainly devoted to finding technical methods to enable computers to intelligently learn and identify real objects in hyperspectral images. Hyperspectral images have great application prospects in many aspects such as urban planning, environmental detection, vegetation classification, military target detection, and mineral geological identification. A general hyperspectral image recognition method is usually: first obtain the spectral features of each pixel from the hyperspectral image, obtain advanced features by extracting the spectral features, and classify the hyperspectral image on this basis. One of the key issues is how to p...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More