CN111076817A - Abnormal target detection method based on optical multi-dimensional information integrated perception system - Google Patents
Abnormal target detection method based on optical multi-dimensional information integrated perception system Download PDFInfo
- Publication number
- CN111076817A CN111076817A CN201911182768.4A CN201911182768A CN111076817A CN 111076817 A CN111076817 A CN 111076817A CN 201911182768 A CN201911182768 A CN 201911182768A CN 111076817 A CN111076817 A CN 111076817A
- Authority
- CN
- China
- Prior art keywords
- polarization
- pixel
- target detection
- information
- abnormal target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 73
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 48
- 230000003287 optical effect Effects 0.000 title claims abstract description 26
- 230000008447 perception Effects 0.000 title claims description 9
- 230000010287 polarization Effects 0.000 claims abstract description 64
- 238000001228 spectrum Methods 0.000 claims abstract description 28
- 238000000034 method Methods 0.000 claims abstract description 24
- 230000003595 spectral effect Effects 0.000 claims abstract description 22
- 238000004590 computer program Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 239000013256 coordination polymer Substances 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 7
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J3/447—Polarisation spectrometry
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Physics & Mathematics (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention relates to an abnormal target detection technology, in particular to an abnormal target detection method based on an optical multi-dimensional information integrated sensing system, which solves the problems of inaccurate abnormal target detection and large spectral data redundancy existing in the conventional abnormal target detection method, aims at the traction of optical remote sensing requirements, aims at extracting more attributes of a target and improving the target detection precision, and mainly comprises the following processes: obtaining polarization spectrum data, reconstructing polarization state information, calculating significance, obtaining a significance map, distributing weight, calculating a detection operator and fusing polarization information.
Description
Technical Field
The invention relates to an abnormal target detection technology, in particular to an abnormal target detection method based on an optical multi-dimensional information integrated perception system.
Background
The optical multi-dimensional information integrated sensing system is a novel optical detection technology which integrates imaging, spectrum, polarization and sensing to acquire multi-dimensional information, and can acquire a data hypercube of a detected target simultaneously, wherein the data hypercube contains two-dimensional space information, spectrum information of each dimension in an image and polarization state information of each spectrum section. The spectral image has the characteristics of high spectral resolution, integrated atlas, multiple spectral channels and the like, and can play great advantages in target detection and identification, so the method has higher application value in the fields of marine ecology, environmental monitoring, military and national defense, urban traffic, accurate agriculture and forestry and the like. The abnormal target detection technology is an unsupervised target detection method, and is used for searching sparse pixels of unknown signals in image data. The technology for detecting the abnormal target by the polarized spectrum is an abnormal target detection technology which is characterized in that the polarized information of the target is added on the basis of a spectrum image and the imaging, the spectrum and the polarized information are integrated, and has important strategic significance in the aspects of novel camouflage disclosure (such as a novel spectrum camouflage technology) and improvement of target identification precision.
At present, a common abnormal target detection method mainly comprises abnormal target detection based on a high spatial resolution image and abnormal target detection based on a hyperspectral image. The abnormal target detection based on the high-spatial-resolution image is mainly implemented by calculating a saliency map by using a saliency detection method in computer vision, and the commonly used methods comprise a sparse and low-rank representation-based method, a Fourier transform-based method, a learning-based method and the like. Abnormal target detection based on the high-spatial-resolution image mainly uses high-resolution spatial data for mining, and the amount of data which can be used by an algorithm is small, so that the abnormal target detection is inaccurate. The abnormal target detection based on the hyperspectral image mainly comprises the steps of searching sparse pixels of unknown spectral signals in the hyperspectral image by utilizing the hyperspectral resolution of the hyperspectral image, and searching pixel points which have obvious differences between spectral characteristics and spectral characteristics of surrounding pixels and appear at low probability. The abnormal target detection based on the hyperspectral image can effectively utilize spectral information and improve the detection precision, but the spectral information is greatly influenced by external factors (such as heavy fog, haze and the like) when being acquired, and the spectral data redundancy is high.
Disclosure of Invention
The invention aims to solve the problems of inaccurate abnormal target detection and high spectral data redundancy existing in the conventional abnormal target detection method, and provides an abnormal target detection method based on an optical multi-dimensional information integrated sensing system. Aiming at the traction required by optical remote sensing, the method aims at extracting more attributes of the target and improving the target detection precision.
In order to achieve the purpose, the invention is realized by the following technical scheme:
an abnormal target detection method based on an optical multi-dimensional information integrated perception system comprises the following steps:
acquiring polarization spectrum data;
an optical multi-dimensional information integrated sensing system acquires polarization spectrum information;
secondly, reconstructing polarization state information;
definition IPM(k) For an input multi-dimensional polarization spectrum image, wherein M is the number of wave bands of the image, k is the number of pixel points, and P is {1, 2, 3, 4}, the polarization state information of different angles is respectively represented, and extracted single-spectrum data I are subjected to single-spectrum data IPM(k)={IP1(k),IP2(k),...,IPM(k) Extracting pixel points in the same polarization direction in the region, and realizing complete reconstruction of spatial information of continuous spectrum segments in different polarization state directions by a neighborhood pixel interpolation method;
step three, calculating significance;
after reconstructing the polarization state information, the significance of the data is calculated, the significance being defined as follows:
wherein i is any one of the local windowsPixel point, j is the center pixel point of the local window, x2(pi,pj) Euclidean distance, x, between the position information in the window for i and j1(pi,pj) Is the spectral distance between i and j, and c is the scale parameter;
step four, obtaining a saliency map;
definition sNThe central pixel point j of the local window is the average value s of the significance of all pixel points i of the local windowNThe calculation method of (2);
wherein, N is the number of all pixel points in the window, a is the number of the pixel points with non-zero significance value in the window; traversing the whole image through a sliding window to calculate s of central pixels of each windowNA saliency map s (k), s (k) ═ s, composing an image1,s2,...,sN};
Step five, calculating the weight;
the weights are calculated as follows:
wherein M represents the wave band number of the image, r represents the spectral curve of the sample pixel, and the sample is assumed to be judged as the background pixel point and obeys the mean value mubCovariance of CbNormal distribution of (1), p (r | H)0) Is the probability that a detected pixel is a background pixel;
normalizing p (r | H) by equation (4)0):
Where N is the number of pixel points, k is the current pixel point, k is 1, 2k|H0) Is the corresponding weight after each pixel is normalized;
the final weight calculation formula is
Wherein, S (k) is the significance map calculated in the step four;
step six, calculating a detection operator;
anomaly detection operator deltaPThe formula (r) is as follows:
δP(r)=(r-μP)TCP -1(r-μP) (6)
wherein the mean vector muPSum covariance matrix CPIs defined as formula (7), (8):
wherein r isiSpectral curve, P, representing the ith pixelP(ri|H0) Is represented by riThe weight of (c);
by traversing the polarization direction P ═ {1, 2, 3, 4}, the anomaly detection operator δ for each polarization direction is calculatedP(r);
Step seven, carrying out polarization information fusion,
the final abnormal target detection operator delta (r) is expressed as follows
Wherein, p is the proportion of each polarization state information;
according to the selected threshold value T, the detection result omega (r) is as follows:
the obtained omega (r) is the detection result, and the pixel point with the value of 1 represents that the abnormal target is detected.
Further, in the first step, the optical multi-dimensional information integrated sensing system acquires polarization spectrum information by using a micro pixilated polarizing film, wherein the pixilated polarizing film is a polarizing film formed by regularly arranging four polarization states, and each polarization state is at a pixel level in physical dimension, and is selected from 0 degree, 45 degrees, 90 degrees and 135 degrees.
Meanwhile, the present invention also provides a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the steps of the abnormal target detection method based on the optical multi-dimensional information-integrated sensing system.
In addition, the invention also provides a computer device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the steps of the abnormal target detection method based on the optical multi-dimensional information integration perception system.
Compared with the prior art, the invention has the following beneficial effects:
1. compared with the traditional high-spatial-resolution target detection, the method can better utilize spectral information and polarization information and fuse the high-spatial-resolution information with the spectral information and the polarization information, so that the target detection effect is better and the detection accuracy of the abnormal target is improved.
2. The method of the invention breaks through the defect that the assumption of multivariate Gaussian distribution of the traditional RX abnormal point detection algorithm cannot completely and truly describe the actual scene, and obviously improves the detection precision.
Drawings
FIG. 1 is a flowchart of an abnormal target detection method of the optical multi-dimensional information integrated sensing system of the present invention;
FIG. 2 is a schematic diagram of polarization state information reconstruction in accordance with the present invention;
FIG. 3a is a schematic diagram of the complete reconstruction of spatial information in the polarization state direction of 0 ° according to the method of the present invention;
FIG. 3b is a schematic diagram of the complete reconstruction of spatial information in the direction of 45 ° polarization according to the method of the present invention;
FIG. 3c is a schematic diagram of the complete reconstruction of spatial information in the 90 ° polarization direction according to the method of the present invention;
FIG. 3d is a schematic diagram of the complete reconstruction of spatial information in the direction of 135 ° polarization state according to the method of the present invention.
Detailed Description
The invention is described in further detail below with reference to the figures and specific embodiments.
The invention provides an abnormal target detection method based on an optical multi-dimensional information integrated sensing system, which is used for detecting an abnormal target by using obtained polarization spectrum data and solves the problems that the available information of the traditional abnormal target detection based on high spatial resolution and the abnormal target detection algorithm based on hyperspectrum is insufficient, and the abnormal target detection result is inaccurate.
As shown in fig. 1, the abnormal target detection method based on the optical multi-dimensional information integrated sensing system provided by the present invention mainly includes the following processes: obtaining polarization spectrum data, reconstructing polarization state information, calculating significance, obtaining a significance map, distributing weight, calculating a detection operator and fusing polarization information.
Acquiring polarization spectrum data;
the optical multidimensional information integrated sensing system adopts a micro pixilated polarizing film to obtain polarization information, the pixilated polarizing film is a polarizing film formed by regularly arranging four polarization states, each polarization state is in a pixel level on the physical dimension, the four polarization states can be flexibly selected, and the device selects 0 degree, 45 degrees, 90 degrees and 135 degrees, and the pixel dimension is 7.4 microns;
secondly, reconstructing polarization state information;
definition IPM(k) For an input multi-dimensional polarization spectrum image, where M is the number of bands, k is the number of pixel points, and P ═ 1, 2, 3, 4, which respectively represent polarization state information of 0 °, 45 °, 90 °, 135 °, the proposed method is to useTaken single spectral band data IPM(k)={IP1(k),IP2(k),...,IPM(k) Extracting pixel points in the same polarization direction in the region, and realizing complete reconstruction of spatial information in different polarization state directions of continuous spectrum segments by a neighborhood pixel interpolation method, as shown in fig. 2;
step three, calculating significance;
after reconstructing the polarization state information, the significance of the data, I, is first calculatedPM(k) For the input multi-dimensional polarization spectrum image, M is the number of wave bands, k is the number of pixel points, P ═ 1, 2, 3, 4}, IPM(k)={IP1(k),IP2(k),...,IPM(k)};
Significance is defined as follows:
wherein, i is any pixel point of the local window, and j is a central pixel point of the local window;
x2(pi,pj) The Euclidean distance between the position information of i and j in the window;
x1(pi,pj) Is the spectral distance between i and j, and c is the scale parameter;
step four, obtaining a saliency map;
definition sNThe central pixel point j of the local window is the average value s of the significance of all pixel points i of the local windowNIs calculated as formula (2);
wherein, N is the number of all pixel points in the window, a is the number of the pixel points with non-zero significance value in the window; traversing the whole image through a sliding window to calculate s of central pixels of each windowNA saliency map s (k), s (k) { s ═ s, composing this image1,s2,...,sN};
Step five, calculating the weight;
the background pixels and the abnormal points are given different weights, so that the influence of the abnormal points and noise is reduced during background model estimation, and the normal adaptability of the background model is better, wherein the weights are calculated as follows:
wherein M represents the wave band number of the image, r represents the spectral curve of the sample pixel, and the sample is assumed to be judged as the background pixel point and obeys the mean value mubCovariance of CbNormal distribution of (1), p (r | H)0) Is the probability that a detected pixel is a background pixel;
normalizing p (r | H) by equation (4)0):
Where N is the number of pixel points, k is the current pixel point, k is 1, 2k|H0) Is the corresponding weight after each pixel is normalized;
the final weight calculation formula is
Wherein, S (k) is the significance map calculated in the step four;
step six, calculating a detection operator;
the detection operator formula is as follows:
δP(r)=(r-μP)TCP -1(r-μP) (6)
δP(r) is the calculated anomaly detection operator, where the mean vector μPSum covariance matrix CPIs defined as formula (7), (8):
wherein r isiSpectral curve, P, representing the ith pixelP(ri|H0) Is represented by riThe weight of (a) is calculated by the formula;
by traversing the polarization direction P ═ {1, 2, 3, 4}, the anomaly detection operator δ for each polarization direction is calculatedP(r);
Step seven, carrying out polarization information fusion, wherein the formula is as follows
Wherein, δ (r) is the final abnormal target detection operator, p is the proportion of each polarization state information, a threshold T is selected according to requirements, and the detection result ω (r) is as follows:
and finally, the obtained omega (r) is a detection result, and the pixel point with the value of 1 represents that an abnormal target is detected.
The method solves the problems of less algorithm available information and inaccurate detection of the existing abnormal target detection method, can obviously increase the algorithm available information, and improves the detection precision.
Compared with the traditional high-spatial-resolution target detection, the method can better utilize spectral information and polarization information and fuse the high-spatial-resolution information with the spectral information and the polarization information, so that the target detection effect is better.
The embodiment of the invention also provides a computer-readable storage medium for storing a program, and the program realizes the steps of the abnormal target detection method based on the optical multi-dimensional information integrated perception system when being executed. In some possible embodiments, the various aspects of the invention may also be implemented in the form of a program product comprising program code means for causing a terminal device to carry out the steps according to various exemplary embodiments of the invention described in the methods presented above in this description, when said program product is run on said terminal device.
In addition, the invention also provides a computer device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein when the processor executes the program, the steps of the abnormal target detection method based on the optical multi-dimensional information integration perception system are realized. A program product for implementing the above method, which may employ a portable compact disc read only memory (CD-ROM) and include program code, may be run on a terminal device, a computer device, such as a personal computer. However, the program product of the present invention is not limited in this regard and, in the present document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
Claims (4)
1. An abnormal target detection method based on an optical multi-dimensional information integrated perception system is characterized by comprising the following steps:
acquiring polarization spectrum data;
an optical multi-dimensional information integrated sensing system acquires polarization spectrum information;
secondly, reconstructing polarization state information;
definition IPM(k) For an input multi-dimensional polarization spectrum image, wherein M is the number of wave bands of the image, k is the number of pixel points, and P is {1, 2, 3, 4}, the polarization state information of different angles is respectively represented, and extracted single-spectrum data I are subjected to single-spectrum data IPM(k)={IP1(k),IP2(k),...,IPM(k) Extracting pixel points in the same polarization direction in the region, and realizing complete reconstruction of spatial information of continuous spectrum segments in different polarization state directions by a neighborhood pixel interpolation method;
step three, calculating significance;
after reconstructing the polarization state information, the significance of the data is calculated, the significance being defined as follows:
wherein, i is any pixel point of the local window, j is the central pixel point of the local window, and x2(pi,pj) Euclidean distance, x, between the position information in the window for i and j1(pi,pj) Is the spectral distance between i and j, and c is the scale parameter;
step four, obtaining a saliency map;
definition sNThe central pixel point j of the local window is the average value s of the significance of all pixel points i of the local windowNIs calculated as formula (2);
wherein, N is the number of all pixel points in the window, a is the number of the pixel points with non-zero significance value in the window; traversing the whole image through a sliding window to calculate s of central pixels of each windowNA saliency map s (k), s (k) being an images1,s2,...,sN};
Step five, calculating the weight;
the weights are calculated as follows:
wherein M represents the wave band number of the image, r represents the spectral curve of the sample pixel, and the sample is assumed to be judged as the background pixel point and obeys the mean value mubCovariance of CbNormal distribution of (1), p (r | H)0) Is the probability that a detected pixel is a background pixel;
normalizing p (r | H) by equation (4)0):
Where N is the number of pixel points, k is the current pixel point, k is 1, 2k|H0) Is the corresponding weight after each pixel is normalized;
the final weight calculation formula is
Step six, calculating a detection operator;
anomaly detection operator deltaPThe formula (r) is as follows:
δP(r)=(r-μP)TCP -1(r-μP) (6)
wherein the mean vector muPSum covariance matrix CPIs defined as formula (7), (8):
wherein r isiSpectral curve, P, representing the ith pixelP(ri|H0) Is represented by riThe weight of (c);
by traversing the polarization direction P ═ {1, 2, 3, 4}, the anomaly detection operator δ for each polarization direction is calculatedP(r);
Step seven, carrying out polarization information fusion,
the final abnormal target detection operator delta (r) is expressed as follows
Wherein, p is the proportion of each polarization state information;
according to the selected threshold value T, the detection result omega (r) is as follows:
the obtained omega (r) is the detection result, and the pixel point with the value of 1 represents that the abnormal target is detected.
2. The abnormal target detection method based on the optical multi-dimensional information integrated perception system according to claim 1, wherein the abnormal target detection method comprises the following steps:
in the first step, the optical multi-dimensional information integrated sensing system adopts a micro pixilated polarizing film to obtain polarization spectrum information, the pixilated polarizing film is a polarizing film formed by regularly arranging four polarization states, and each polarization state is in a pixel level on the physical dimension and is selected from 0 degree, 45 degrees, 90 degrees and 135 degrees.
3. A computer-readable storage medium having stored thereon a computer program, characterized in that: which computer program, when being executed by a processor, carries out the steps of the method as set forth in claim 1 or 2.
4. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein: the processor, when executing the program, implements the steps of the method of claim 1 or 2.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911182768.4A CN111076817B (en) | 2019-11-27 | 2019-11-27 | Abnormal target detection method based on optical multi-dimensional information integrated perception system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911182768.4A CN111076817B (en) | 2019-11-27 | 2019-11-27 | Abnormal target detection method based on optical multi-dimensional information integrated perception system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111076817A true CN111076817A (en) | 2020-04-28 |
CN111076817B CN111076817B (en) | 2021-04-20 |
Family
ID=70311859
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911182768.4A Active CN111076817B (en) | 2019-11-27 | 2019-11-27 | Abnormal target detection method based on optical multi-dimensional information integrated perception system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111076817B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112488939A (en) * | 2020-11-27 | 2021-03-12 | 深圳市中博科创信息技术有限公司 | Image processing method, terminal device and storage medium |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170148166A1 (en) * | 2014-02-12 | 2017-05-25 | International Business Machines Corporation | Anomaly detection in medical imagery |
-
2019
- 2019-11-27 CN CN201911182768.4A patent/CN111076817B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170148166A1 (en) * | 2014-02-12 | 2017-05-25 | International Business Machines Corporation | Anomaly detection in medical imagery |
Non-Patent Citations (1)
Title |
---|
刘嘉诚 等: "显著性权重RX高光谱异常点检测", 《遥感学报》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112488939A (en) * | 2020-11-27 | 2021-03-12 | 深圳市中博科创信息技术有限公司 | Image processing method, terminal device and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN111076817B (en) | 2021-04-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Qiu et al. | Deeplidar: Deep surface normal guided depth prediction for outdoor scene from sparse lidar data and single color image | |
Xia et al. | Infrared small target detection based on multiscale local contrast measure using local energy factor | |
CN111222395B (en) | Target detection method and device and electronic equipment | |
US20180189577A1 (en) | Systems and methods for lane-marker detection | |
CN111680579B (en) | Remote sensing image classification method for self-adaptive weight multi-view measurement learning | |
CN110555339A (en) | target detection method, system, device and storage medium | |
CN115131580B (en) | Space target small sample identification method based on attention mechanism | |
CN110443169B (en) | Face recognition method based on edge preservation discriminant analysis | |
CN111076817B (en) | Abnormal target detection method based on optical multi-dimensional information integrated perception system | |
Messmer et al. | Gaining scale invariance in uav bird’s eye view object detection by adaptive resizing | |
CN113446998B (en) | Hyperspectral target detection data-based dynamic unmixing method | |
US8126275B2 (en) | Interest point detection | |
JP2015092339A (en) | Method of identification from spatial and spectral object model | |
Zhu et al. | Unmanned aerial vehicle remote sensing image registration based on an improved oriented FAST and rotated BRIEF-random sample consensus algorithm | |
Li et al. | Multi-scale ships detection in high-resolution remote sensing image via saliency-based region convolutional neural network | |
CN109035199B (en) | Hyperspectral data end member extraction method, storage medium and electronic device | |
WO2017166156A1 (en) | Hyperspectral image classification method and system | |
CN116953702A (en) | Rotary target detection method and device based on deduction paradigm | |
CN113591890A (en) | Clustering method and device | |
CN111582013A (en) | Ship retrieval method and device based on gray level co-occurrence matrix characteristics | |
CN111652034A (en) | Ship retrieval method and device based on SIFT algorithm | |
US20240127585A1 (en) | Method for determining the encoder architecture of a neural network | |
Guo et al. | Novel registration and fusion algorithm for multimodal railway images with different field of views | |
CN106033545B (en) | Wave band selection method of determinant point process | |
Li et al. | A new framework of hyperspectral image classification based on spatial spectral interest point |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |