A Progressive Dynamic Hyperspectral Image Classification Method
A hyperspectral image and classification method technology, applied in the field of progressive dynamic hyperspectral image classification, can solve problems such as little thinking
Active Publication Date: 2021-09-28
江门市华讯方舟科技有限公司
View PDF7 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, the existing HIC methods either focus on the use of classifiers or on feature extraction methods, but there is little thought. If HIC needs manual classification, how to make full use of the local continuity characteristics of the distribution of ground objects and learn from human Processes and strategies for manual HIC to minimize classification errors and establish efficient HIC methods
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
[0036] The hyperspectral image used in this example is AVIRIS Indian pins. The size of the hyperspectral image is 145×145, with 220 spectral segments, uniformly covering the spectral range of 0.2-2.5 μm. The hyperspectral image contains 16 types of labeled samples. Due to absorption by water and low signal-to-noise ratio, bands 104-108, 150-163, and 220 were removed before classification, leaving only a total of 200 bands. figure 2 A false-color image of the AVIRIS Indian pins is given.
[0037] refer to figure 1 , the specific implementation steps of a progressive dynamic hyperspectral image classification method are as follows:
[0038] Step 1: Input hyperspectral image data D∈R 145×145×200 And the corresponding feature marker matrix L∈R 145×145 , each pixel or sample in D is represented by a hyperspectral feature vector, and the dimension of the sample is 200; L(x, y)=c indicates that the pixel at the image position (x, y) belongs to the c-th class (c=1, 2...16); ran...
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 a progressive dynamic hyperspectral image classification method, which comprises the following steps: step 1: reading hyperspectral image data; step 2: calculating the difference between each pixel in the hyperspectral image and its local neighborhood pixels, by setting Determine a suitable threshold, perform discontinuity detection, and classify continuous points and discontinuous points; Step 3, by comparing the continuous points with the points of known categories in their local neighborhoods, cyclically classify the unclassified continuous points; Step 4, cyclically classify discontinuous points and continuous points that cannot be classified in Step 3; Step 5, classify all remaining points to be classified in Step 4; Step 6, output classification results. This patent proposes a progressive dynamic hyperspectral classification method. Compared with most existing methods, this method does not require advanced mathematical knowledge, and has the advantages of low computational complexity, high classification accuracy, and fast operation speed.
Description
technical field [0001] The invention relates to the technical field of hyperspectral image classification, in particular to a progressive dynamic hyperspectral image classification method. Background technique [0002] After decades of development, hyperspectral classification (HIC) methods have been very rich. Early HIC methods focused on directly borrowing classifiers commonly used in other fields for HIC. Such as methods based on classifiers such as SVM and decision trees. Later, according to the particularity of HIC, people began to integrate spatial information into HIC, and proposed a spectral-spatial HIC method. The spectral-spatial HIC method is currently the most mainstream HIC method. Among them, methods based on superpixel segmentation have gained more attention in recent years. With the development of deep learning in various fields, HIC based on deep learning has also begun to appear. [0003] The reason why human beings are different from other animals is ...
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): G06K9/62
CPCG06F18/24
Inventor 郑成勇
Owner 江门市华讯方舟科技有限公司