CN112418314A - Threshold setting method and device in spectrum similarity matching classification - Google Patents
Threshold setting method and device in spectrum similarity matching classification Download PDFInfo
- Publication number
- CN112418314A CN112418314A CN202011322045.2A CN202011322045A CN112418314A CN 112418314 A CN112418314 A CN 112418314A CN 202011322045 A CN202011322045 A CN 202011322045A CN 112418314 A CN112418314 A CN 112418314A
- Authority
- CN
- China
- Prior art keywords
- similarity
- threshold
- actual
- value
- variance
- 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
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000001228 spectrum Methods 0.000 title claims description 49
- 230000003595 spectral effect Effects 0.000 claims abstract description 68
- 238000010586 diagram Methods 0.000 claims description 12
- 238000011156 evaluation Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 abstract description 5
- 238000001914 filtration Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 description 6
- 230000004044 response Effects 0.000 description 4
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Probability & Statistics with Applications (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to the technical field of image data processing, and particularly relates to a threshold setting method and device in spectral similarity matching classification, wherein the method comprises the steps of obtaining a target region of interest selected from a hyperspectral image of a preselected object; calculating actual similarity values of the spectral data of each pixel in the target region of interest and standard spectral data of each category in a preset spectral library; calculating the mean and variance of each actual similarity value; and obtaining the following threshold setting range according to the mean value and the variance of each actual similarity value. The invention realizes that reliable data reference is provided for threshold setting according to the mean value and the variance, thereby improving the threshold adjusting efficiency, filtering out setting steps of useless and irrelevant thresholds, simplifying the threshold setting process, improving the classification efficiency and having extremely high practicability.
Description
Technical Field
The invention belongs to the technical field of image data processing, and particularly relates to a threshold setting method and device in spectral similarity matching classification.
Background
The classification of objects, such as the classification of garbage, is a technology that is continuously sought and updated by the market. The classification method commonly used in the market is, for example, classification by using spectral similarity, and classification by using spectral similarity.
When the classification is performed by using the spectrum similarity, a similarity comparison threshold needs to be set, and the classification category to which the classification belongs is determined by the similarity threshold. At present, technical personnel are purposeless setting when setting a threshold value, detect the classification effect after the threshold value is set one by one, and then adjust according to the classification effect, so, because the threshold value is set without data reference, the final threshold value can be set through many threshold value adjustments, which leads to the technical problems that the threshold value setting process is complicated and the classification efficiency is affected. Therefore, it is necessary to design a method and an apparatus for setting a threshold in a spectral similarity matching classification.
Disclosure of Invention
The invention aims to provide a threshold setting method and a threshold setting device in spectral similarity matching classification, and aims to solve the technical problems that in the prior art, when classification is carried out through spectral similarity matching, the threshold setting process is complicated and the classification efficiency is influenced because no data reference exists in the set threshold.
In order to achieve the above object, an embodiment of the present invention provides a threshold setting method in spectrum similarity matching classification, where the method includes:
acquiring a target region of interest selected from a hyperspectral image of a preselected object;
calculating actual similarity values of the spectral data of each pixel in the target region of interest and standard spectral data of each category in a preset spectral library;
calculating the mean and variance of each actual similarity value;
obtaining the following threshold setting range according to the mean value and the variance of each actual similarity value:
[μ-δ,μ+δ]；
where μ is the actual mean and δ is the actual variance.
Optionally, the calculating a mean and a variance of each of the actual similarity values further includes:
and generating a spectrum classification evaluation graph according to the mean value and the variance of the actual similarity value.
Optionally, after obtaining the following threshold setting range according to the mean and the variance of each of the actual similarity values, the method further includes:
acquiring similarity threshold values of various types of spectrums in a spectrum library respectively set according to the threshold setting range; wherein a spectral class has a similarity threshold;
calculating the percentage of the number of pixels corresponding to the actual similarity value meeting the similarity threshold value in the total number of the corresponding class pixels in the region of interest, and recording the percentage as the known classification percentage;
and generating a threshold setting performance indication diagram according to the known classification percentage, wherein the threshold setting performance indication diagram is marked with the known classification percentage corresponding to each category.
Optionally, the calculating a percentage of the number of pixels corresponding to the actual similarity value satisfying the similarity threshold to the total number of corresponding class pixels in the region of interest, and the recording as the known classification percentage includes:
comparing each actual similarity value with each similarity threshold value respectively;
counting the number of pixels corresponding to the actual similarity value meeting the similarity threshold value, and recording as known classification pixels;
and calculating the percentage of the known classification pixels to the total number of the corresponding class pixels in the region of interest, and recording the percentage as the known classification percentage.
Optionally, the acquiring a selected target region of interest from a hyperspectral image of a preselected object comprises:
acquiring a hyperspectral image of a preselected object;
and selecting a target region of interest in the hyperspectral image.
Optionally, there is further provided a threshold setting device in the spectral similarity matching classification, the device including:
the interesting region acquisition module is used for acquiring a target interesting region selected from a hyperspectral image of a preselected object;
the actual similarity calculation module is used for calculating the actual similarity value between the spectral data of each pixel in the target region of interest and the standard spectral data of each category in a preset spectral library;
the mean value and variance calculation module is used for calculating the mean value and variance of each actual similarity value;
a threshold setting range obtaining module, configured to obtain the following threshold setting ranges according to the mean and variance of each of the actual similarity values:
[μ-δ,μ+δ]；
where μ is the actual mean and δ is the actual variance.
Optionally, the apparatus further comprises an evaluation graph generation module for generating a spectral classification evaluation graph according to the mean and variance of the actual similarity values.
Optionally, the apparatus comprises:
the similarity threshold acquisition module is used for acquiring similarity thresholds of various types of spectrums in the spectrum library respectively set according to the threshold setting range; wherein a spectral class has a similarity threshold;
a known classification percentage calculation module, configured to calculate a percentage of the number of pixels corresponding to the actual similarity value that meets the similarity threshold to the total number of corresponding class pixels in the region of interest, and record the percentage as a known classification percentage;
and the performance indication map generation module is used for generating a threshold setting performance indication map according to the known classification percentage, and the threshold setting performance indication map marks the known classification percentage corresponding to each class.
Optionally, the present invention further provides a computer device, which includes a memory and a processor, where the memory stores a computer program, and the processor implements the steps in the threshold setting method in the spectral similarity matching classification when executing the computer program.
Optionally, the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps in the above-mentioned threshold setting method in the spectral similarity matching classification.
The one or more technical solutions in the method and the device for setting the threshold in the spectral similarity matching classification provided by the embodiment of the present invention at least have one of the following technical effects:
according to the method, a target interesting area selected from a hyperspectral image of a preselected object is obtained, actual similarity values of spectral data of each pixel in the target interesting area and standard spectral data of various types in a preset spectrum library are calculated, then the mean value and the variance of each actual similarity value are calculated, and a threshold setting range is set according to the mean value and the variance, so that reliable data reference is provided for threshold setting according to the mean value and the variance, threshold adjusting efficiency is improved, setting steps of filtering useless and irrelevant thresholds are eliminated, a threshold setting process is simplified, classification efficiency is improved, and the method has extremely high practicability.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise.
Fig. 1 is a flowchart of steps S100-S400 in a method for setting a threshold in a spectral similarity matching classification according to an embodiment of the present invention;
FIG. 2 is a diagram of spectral classification evaluation of the mean and variance of the actual similarity values provided by an embodiment of the present invention;
FIG. 3 is a flowchart illustrating steps S510-S530 of a threshold setting method for spectral similarity matching classification according to an embodiment of the present invention;
FIG. 4 is a diagram of threshold setting performance indicators according to an embodiment of the present invention;
fig. 5 is a flowchart of steps S531-S533 in the method for setting a threshold in spectral similarity matching classification according to the embodiment of the present invention;
FIG. 6 is a flowchart illustrating steps S331-S332 of the method for setting threshold values in spectral similarity matching classification according to an embodiment of the present invention;
fig. 7 is a block diagram of a threshold setting apparatus in the spectrum similarity matching classification according to an embodiment of the present invention;
FIG. 8 is a block diagram of a computer device according to an embodiment of the present invention;
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It should also be understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
As used in this specification and the appended claims, the term "if" may be interpreted contextually as "when", "upon" or "in response to" determining "or" in response to detecting ". Similarly, the phrase "if it is determined" or "if a [ described condition or event ] is detected" may be interpreted contextually to mean "upon determining" or "in response to determining" or "upon detecting [ described condition or event ]" or "in response to detecting [ described condition or event ]".
Furthermore, in the description of the present application and the appended claims, the terms "first," "second," "third," and the like are used for distinguishing between descriptions and not necessarily for describing or implying relative importance.
Reference throughout this specification to "one embodiment" or "some embodiments," or the like, means that a particular feature, structure, or characteristic described in connection with the embodiment is included in one or more embodiments of the present application. Thus, appearances of the phrases "in one embodiment," "in some embodiments," "in other embodiments," or the like, in various places throughout this specification are not necessarily all referring to the same embodiment, but rather "one or more but not all embodiments" unless specifically stated otherwise. The terms "comprising," "including," "having," and variations thereof mean "including, but not limited to," unless expressly specified otherwise.
In an embodiment of the present invention, as shown in fig. 1, a method for setting a threshold in a spectral similarity matching classification is provided, the method including the following steps:
step S100: acquiring a target region of interest selected from a hyperspectral image of a preselected object;
in this step, the preselected object is selected by one skilled in the art. The number of the preselected objects may be one or more, and in this embodiment, the number of the preselected objects is more than one.
Furthermore, for the convenience of threshold value adjustment, in this embodiment, a plurality of hyperspectral images of the preselected object are collected in the same hyperspectral image, and when a target region of interest is collected, spectral data of a plurality of objects are contained in the target region of interest, and only one hyperspectral image needs to be processed during subsequent classification, so that the processing efficiency is improved.
Step S200: calculating actual similarity values of the spectral data of each pixel in the target region of interest and standard spectral data of each category in a preset spectral library;
in this step, the spectrum library includes a plurality of categories, and each category corresponds to standard spectrum data. The pixels in the target region of interest also have multiple categories. During classification, the spectral data of each pixel in the target region of interest needs to be compared with each standard spectral data one by one, and the actual similarity of the two compared spectral data is calculated.
Step S300: calculating the mean and variance of each actual similarity value;
specifically, three categories in the spectral library are exemplified:
specifically, the three categories in the spectral library are a category, B category, and C category, respectively. The A category, the B category and the C category are respectively corresponding to standard spectrum data.
The target region of interest also has known pixels belonging to class a, class B and class C.
Specifically, a pixel is selected from the target region of interest, and the pixel is compared with the standard spectrum data corresponding to the type a in the spectrum library, so that the actual similarity value of the pixel can be obtained. Next, the mean and variance of the actual similarity values obtained above are calculated, where the mean is denoted as μ a and the variance is denoted as δ a.
Similarly, the variance and mean of the actual similarity values after comparison with the B and C categories are calculated respectively. Wherein, the mean value corresponding to the B category is recorded as μ B, the variance is recorded as δ B, the mean value corresponding to the C category is recorded as μ C, and the variance is recorded as δ C.
That is, in this step, the mean value is the mean value of the actual similarity obtained after the pixels in the target region of interest are compared with the same category in the spectrum library. The variance is the same.
Step S400: obtaining the following threshold setting range according to the mean value and the variance of each actual similarity value:
[μ-δ,μ+δ]；
where μ is the actual mean and δ is the actual variance.
In this step, the threshold setting range obtained by the mean and the variance provides an initial threshold setting reference range for the user, and compared with the threshold which is set randomly without purpose by the user, the threshold is adjusted, so that the threshold setting is accurately set, and the threshold setting efficiency is improved.
According to the method, a target interesting area selected from a hyperspectral image of a preselected object is obtained, actual similarity values of spectral data of each pixel in the target interesting area and standard spectral data of various types in a preset spectrum library are calculated, then the mean value and the variance of each actual similarity value are calculated, and a threshold setting range is set according to the mean value and the variance, so that reliable data reference is provided for threshold setting according to the mean value and the variance, threshold adjusting efficiency is improved, setting steps of filtering useless and irrelevant thresholds are eliminated, a threshold setting process is simplified, classification efficiency is improved, and the method has extremely high practicability.
In another embodiment of the present invention, as shown in fig. 2, after the calculating the mean and the variance of each of the actual similarity values, the method further includes:
and generating a spectrum classification evaluation graph according to the mean value and the variance of the actual similarity value.
In this step, the mean and the variance are obtained by referring to the above example.
And drawing a circle by taking the mean value as a circle center and the variance as a radius to generate the spectrum classification evaluation graph.
As shown in fig. 2, where the ordinate is the class classification in the spectral library. The circle center of the A-A is the mean value of the actual similarity between the A-type spectrum in the target region of interest and the pixels corresponding to all the A types in the spectrum library, and the radius of the A-A is the variance of the actual similarity between the A-type spectrum in the target region of interest and the pixels corresponding to all the A types in the spectrum library.
The circle center of the A-B is the mean value of the actual similarity of the A-type spectrum in the target region of interest and the pixels corresponding to all the B types in the spectrum library, and the radius of the A-B is the variance of the actual similarity of the A-type spectrum in the target region of interest and the pixels corresponding to all the B types in the spectrum library.
The circle center of the A-C is the mean value of the actual similarity of the A-type spectrum in the target region of interest and the pixels corresponding to all the C types in the spectrum library, and the radius of the A-C is the variance of the actual similarity of the C-type spectrum in the target region of interest and the pixels corresponding to all the C types in the spectrum library.
Similarly, the circle center of the B-B is the mean value of the actual similarity between the B-type spectrum in the target region of interest and the pixels corresponding to all the B types in the spectrum library, and the radius of the B-B is the variance of the actual similarity between the B-type spectrum in the target region of interest and the pixels corresponding to all the B types in the spectrum library. Similarly, B-A, B-C, C-C, C-B and C-A are known.
For A-A, A-B and A-C, when the circle center of A-A is closer to the left side and the radius of A-A is smaller, the fact that the actual similarity of the same category is small in fluctuation and the classification effect is good is shown. When the distances between A-A and A-B and A-C are farther and farther, the similarity difference of different categories is very large, and the classification effect is good.
Therefore, the classification effect can be visually seen through the spectrum classification evaluation graph.
In another embodiment of the present invention, as shown in fig. 3, after obtaining the following threshold setting ranges according to the mean and variance of each of the actual similarity values, the method further includes:
step S510: acquiring similarity threshold values of various types of spectrums in a spectrum library respectively set according to the threshold setting range; wherein a spectral class has a similarity threshold;
in this step, when the initial reference range set by the threshold is obtained, a similarity threshold may be set for each category in the spectral library. As in the above embodiment, the spectrum library includes A, B and C, so that the three categories can be set with corresponding similarity thresholds.
Step S520: calculating the percentage of the number of pixels corresponding to the actual similarity value meeting the similarity threshold value in the total number of the corresponding class pixels in the region of interest, and recording the percentage as the known classification percentage;
specifically, if there are 1000 pixels in the class a in the target region of interest, after actual comparison, the number of pixels corresponding to the actual similarity value satisfying the similarity threshold is 800, so the known classification percentage is 80%. 80% is the known classification percentage corresponding to the pixels in class A, and other classes can be obtained by the same method.
Step S530: and generating a threshold setting performance indication diagram according to the known classification percentage, wherein the threshold setting performance indication diagram is marked with the known classification percentage corresponding to each category.
Specifically, in this step, the threshold setting performance indicator is formed by collecting the known classification percentages of the classes in the same graph, as shown in FIG. 4, where A-A, B-B and C-C are labeled A, B and C, respectively.
Therefore, through the threshold setting performance indication diagram, when the threshold of each category is modified respectively, different threshold setting performance indication diagrams can be formed, and when the correct percentage of each category is close to 1, the threshold of each category is the target threshold, so that visual efficient threshold adjustment is realized, the threshold is quickly set, and the threshold setting efficiency is improved.
In another embodiment of the present invention, as shown in fig. 5, the calculating the percentage of the number of pixels corresponding to the actual similarity value satisfying the similarity threshold to the total number of pixels of the corresponding category in the region of interest, and the recording as the known classification percentage includes:
step S531: comparing each actual similarity value with each similarity threshold value respectively;
step S532: counting the number of pixels corresponding to the actual similarity value meeting the similarity threshold value, and recording as known classification pixels;
step S533: and calculating the percentage of the known classification pixels to the total number of the corresponding class pixels in the region of interest, and recording the percentage as the known classification percentage.
Specifically, the known classification percentage can be calculated through the steps S531-S533 conveniently and quickly.
In another embodiment of the present invention, as shown in fig. 6, the acquiring a target region of interest selected from a hyperspectral image of a preselected object comprises:
step S110: acquiring a hyperspectral image of a preselected object;
in this step, the hyperspectral image is acquired by a hyperspectral camera.
Step S120: and selecting a target region of interest in the hyperspectral image.
Specifically, after the target region of interest is selected, the subsequent data emphasis processing in the target region of interest is realized, and the processing efficiency is improved.
In another embodiment of the present invention, as shown in fig. 7, there is further provided a threshold setting apparatus in spectral similarity matching classification, the apparatus including a region of interest obtaining module, an actual similarity calculating module, a mean and variance calculating module, and a threshold setting range obtaining module.
The interesting region acquisition module is used for acquiring a target interesting region selected from a hyperspectral image of a preselected object;
the actual similarity calculation module is used for calculating the actual similarity value between the spectral data of each pixel in the target region of interest and the standard spectral data of each category in the preset spectral library;
the mean value and variance calculation module is used for calculating the mean value and variance of each actual similarity value;
the threshold setting range obtaining module is configured to obtain the following threshold setting ranges according to the mean and variance of each actual similarity value:
[μ-δ,μ+δ]；
where μ is the actual mean and δ is the actual variance.
In another embodiment of the present invention, the apparatus for setting a threshold in spectral similarity matching classification further includes an evaluation graph generating module, configured to generate a spectral classification evaluation graph according to a mean and a variance of the actual similarity value.
In another embodiment of the present invention, the apparatus for setting a threshold in spectral similarity matching classification further includes a similarity threshold obtaining module, a known classification percentage calculating module, and a performance indicator map generating module.
The similarity threshold acquisition module is used for acquiring similarity thresholds of various types of spectrums in the spectrum library respectively set according to the threshold setting range; wherein a spectral class has a similarity threshold;
the known classification percentage calculation module is used for calculating the percentage of the number of pixels corresponding to the actual similarity value meeting the similarity threshold value in the total number of the corresponding class pixels in the region of interest and recording the percentage as the known classification percentage;
the performance indication map generation module is configured to generate a threshold setting performance indication map according to the known classification percentage, where the threshold setting performance indication map indicates the known classification percentage corresponding to each class.
In another embodiment of the present invention, as shown in fig. 8, there is further provided a computer device, including a memory and a processor, where the memory stores a computer program, and the processor implements the steps in the threshold setting method in the above-mentioned spectral similarity matching classification when executing the computer program.
In another embodiment of the present invention, a computer-readable storage medium is further provided, on which a computer program is stored, which when executed by a processor implements the steps in the above-mentioned threshold setting method in the spectral similarity matching classification.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. A threshold setting method in spectral similarity matching classification is characterized by comprising the following steps:
acquiring a target region of interest selected from a hyperspectral image of a preselected object;
calculating actual similarity values of the spectral data of each pixel in the target region of interest and standard spectral data of each category in a preset spectral library;
calculating the mean and variance of each actual similarity value;
obtaining the following threshold setting range according to the mean value and the variance of each actual similarity value:
[μ-δ,μ+δ]；
where μ is the actual mean and δ is the actual variance.
2. The method as claimed in claim 1, wherein the step of calculating the mean and variance of each actual similarity value further comprises:
and generating a spectrum classification evaluation graph according to the mean value and the variance of the actual similarity value.
3. The method for setting the threshold value in the spectral similarity matching classification according to claim 1, wherein after obtaining the following threshold setting ranges according to the mean and variance of each actual similarity value, the method further comprises:
acquiring similarity threshold values of various types of spectrums in a spectrum library respectively set according to the threshold setting range; wherein a spectral class has a similarity threshold;
calculating the percentage of the number of pixels corresponding to the actual similarity value meeting the similarity threshold value in the total number of the corresponding class pixels in the region of interest, and recording the percentage as the known classification percentage;
and generating a threshold setting performance indication diagram according to the known classification percentage, wherein the threshold setting performance indication diagram is marked with the known classification percentage corresponding to each category.
4. The method as claimed in claim 3, wherein the calculating the percentage of the number of pixels corresponding to the actual similarity value satisfying the similarity threshold to the total number of pixels corresponding to the category in the region of interest and recording as the known classification percentage includes:
comparing each actual similarity value with each similarity threshold value respectively;
counting the number of pixels corresponding to the actual similarity value meeting the similarity threshold value, and recording as known classification pixels;
and calculating the percentage of the known classification pixels to the total number of the corresponding class pixels in the region of interest, and recording the percentage as the known classification percentage.
5. The method for setting the threshold value in the spectral similarity matching classification according to claim 1, wherein the acquiring the target region of interest selected from the hyperspectral image of the preselected object comprises:
acquiring a hyperspectral image of a preselected object;
and selecting a target region of interest in the hyperspectral image.
6. An apparatus for threshold setting in spectral similarity matching classification, the apparatus comprising:
the interesting region acquisition module is used for acquiring a target interesting region selected from a hyperspectral image of a preselected object;
the actual similarity calculation module is used for calculating the actual similarity value between the spectral data of each pixel in the target region of interest and the standard spectral data of each category in a preset spectral library;
the mean value and variance calculation module is used for calculating the mean value and variance of each actual similarity value;
a threshold setting range obtaining module, configured to obtain the following threshold setting ranges according to the mean and variance of each of the actual similarity values:
[μ-δ,μ+δ]；
where μ is the actual mean and δ is the actual variance.
7. The apparatus for setting the threshold value in the spectrum similarity matching classification as claimed in claim 6, further comprising an evaluation graph generating module for generating the spectrum classification evaluation graph according to the mean and variance of the actual similarity value.
8. The apparatus for thresholding in spectral similarity matching classification according to claim 6, further comprising:
the similarity threshold acquisition module is used for acquiring similarity thresholds of various types of spectrums in the spectrum library respectively set according to the threshold setting range; wherein a spectral class has a similarity threshold;
a known classification percentage calculation module, configured to calculate a percentage of the number of pixels corresponding to the actual similarity value that meets the similarity threshold to the total number of corresponding class pixels in the region of interest, and record the percentage as a known classification percentage;
and the performance indication map generation module is used for generating a threshold setting performance indication map according to the known classification percentage, and the threshold setting performance indication map marks the known classification percentage corresponding to each class.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 5 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011322045.2A CN112418314B (en) | 2020-11-23 | 2020-11-23 | Method and device for setting threshold value in spectrum similarity matching classification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011322045.2A CN112418314B (en) | 2020-11-23 | 2020-11-23 | Method and device for setting threshold value in spectrum similarity matching classification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112418314A true CN112418314A (en) | 2021-02-26 |
| CN112418314B CN112418314B (en) | 2024-06-28 |
Family
ID=74777920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011322045.2A Active CN112418314B (en) | 2020-11-23 | 2020-11-23 | Method and device for setting threshold value in spectrum similarity matching classification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112418314B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023092707A1 (en) * | 2021-11-25 | 2023-06-01 | 中国科学院深圳先进技术研究院 | Multispectral image generation method, terminal device and computer-readable storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101950361A (en) * | 2010-09-06 | 2011-01-19 | 中国科学院遥感应用研究所 | Adaptive extraction method of remote sensing image thematic information based on spectrum matching degree |
| WO2018076138A1 (en) * | 2016-10-24 | 2018-05-03 | 深圳大学 | Target detection method and apparatus based on large-scale high-resolution hyper-spectral image |
| US20180365820A1 (en) * | 2017-06-19 | 2018-12-20 | ImpactVision, Inc. | System and method for hyperspectral image processing to identify object |
| CN110288647A (en) * | 2019-06-25 | 2019-09-27 | 中国水利水电科学研究院 | One kind monitoring Irrigation Project Design area method based on high resolution satellite remote sensing image |
-
2020
- 2020-11-23 CN CN202011322045.2A patent/CN112418314B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101950361A (en) * | 2010-09-06 | 2011-01-19 | 中国科学院遥感应用研究所 | Adaptive extraction method of remote sensing image thematic information based on spectrum matching degree |
| WO2018076138A1 (en) * | 2016-10-24 | 2018-05-03 | 深圳大学 | Target detection method and apparatus based on large-scale high-resolution hyper-spectral image |
| US20180365820A1 (en) * | 2017-06-19 | 2018-12-20 | ImpactVision, Inc. | System and method for hyperspectral image processing to identify object |
| CN110288647A (en) * | 2019-06-25 | 2019-09-27 | 中国水利水电科学研究院 | One kind monitoring Irrigation Project Design area method based on high resolution satellite remote sensing image |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023092707A1 (en) * | 2021-11-25 | 2023-06-01 | 中国科学院深圳先进技术研究院 | Multispectral image generation method, terminal device and computer-readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112418314B (en) | 2024-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11544628B2 (en) | Information processing apparatus and information processing method for generating classifier using target task learning data and source task learning data, and storage medium | |
| CN108932496B (en) | Method and device for counting number of target objects in area | |
| US7054489B2 (en) | Apparatus and method for image recognition | |
| US7783106B2 (en) | Video segmentation combining similarity analysis and classification | |
| US11194073B2 (en) | Millimeter wave image based human body foreign object detection method and system | |
| US10529103B2 (en) | Image processing apparatus and method for collating a plurality of images | |
| CN105160355B (en) | A kind of method for detecting change of remote sensing image based on region correlation and vision word | |
| CN116091883B (en) | Target detection and identification method, system and storage medium based on multi-source information fusion | |
| Short et al. | Latent fingerprint segmentation using ridge template correlation | |
| WO2019076187A1 (en) | Video blocking region selection method and apparatus, electronic device, and system | |
| CN110889355A (en) | Face recognition verification method, system and storage medium | |
| WO2019102608A1 (en) | Image processing device, image processing method, and image processing program | |
| CN111814612A (en) | Target face detection method and related device thereof | |
| CN110619625B (en) | Method, device and system for monitoring running state of belt and storage medium | |
| CN112418314A (en) | Threshold setting method and device in spectrum similarity matching classification | |
| CN111814659B (en) | Living body detection method and system | |
| JP2008165496A (en) | Image normalization device, object detection device, object detection system and program | |
| CN115050069A (en) | Face and attribute recognition method and system based on deep learning and computer equipment | |
| DE102015200434A1 (en) | Method and apparatus for improving object recognition in different lighting situations | |
| JP6457728B2 (en) | Laminar smoke detection device and laminar smoke detection method | |
| US20240303828A1 (en) | Device and method for enhancing tracking of objects in a scene captured in a video sequence | |
| CN107330105B (en) | Robustness evaluation method and device for similar image retrieval algorithm | |
| DE102015200437A1 (en) | Method and device for determining the confidence of an object recognition | |
| JP6457727B2 (en) | Laminar smoke detection device and laminar smoke detection method | |
| CN112418104B (en) | Pedestrian tracking method and related equipment |
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 |