CN107067397A - A kind of image grading method based on infrared image complexity - Google Patents
A kind of image grading method based on infrared image complexity Download PDFInfo
- Publication number
- CN107067397A CN107067397A CN201710289572.XA CN201710289572A CN107067397A CN 107067397 A CN107067397 A CN 107067397A CN 201710289572 A CN201710289572 A CN 201710289572A CN 107067397 A CN107067397 A CN 107067397A
- Authority
- CN
- China
- Prior art keywords
- image
- mrow
- complexity
- munderover
- msub
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention provides a kind of image grading method based on infrared image complexity.This method includes：The selected sample image sequence that need to be measured, is used as sample data set；The every piece image concentrated for sample data, calculates the characteristics of image of the image, and the characteristics of image obtained according to calculating, calculates the image complexity for obtaining image respectively；Z width images are selected, the image complexity of selected Z width images is counted, complexity statistical result is obtained；The every piece image concentrated for sample data, the rank of the image is determined according to the image complexity of the image and the complexity statistical result.Image grading can be carried out according to image complexity using the present invention.
Description
Technical field
The application is related to technical field of image processing, more particularly to a kind of image grading side based on infrared image complexity
Method.
Background technology
In the prior art, infrared image amount of complexity method can be not only used for describing the first-class infrared imaging of infrared seeker
The complexity of the working environment of system, and in Infrared images pre-processing algorithms selection and Performance Evaluation, infrared imaging system
It can predict and assess, Target Recognition Algorithms performance comparison, foundation and Further aim have also obtained wide in terms of obtaining performance model
General and important application.
However, in the prior art, the computational methods on image complexity, there is no unified calculating standard at present, and
It is each, according to different application, to define its calculation formula, therefore can not unify and accurately carry out image grading.
The content of the invention
In view of this, the invention provides a kind of image grading method based on infrared image complexity, so as to root
Image grading is carried out according to image complexity.
What technical scheme was specifically realized in：
A kind of image grading method based on infrared image complexity, this method includes：
The selected sample image sequence that need to be measured, is used as sample data set；
The every piece image concentrated for sample data, calculates the characteristics of image of the image, and obtain according to calculating respectively
Characteristics of image, calculate and obtain the image complexity of image；
Z width images are selected, the image complexity of selected Z width images is counted, complexity statistical result is obtained；
The every piece image concentrated for sample data, is counted according to the image complexity of the image and the complexity
As a result the rank of the image is determined.
Preferably, described image feature includes：
Image information entropy, image border ratio, image contrast, image correlativity and image energy.
Preferably, calculating the image complexity of piece image by formula below：
Wherein, C is the image complexity of image, and k is the number of different images feature in sample data set, and n is that image is special
Levy, a is the individual amount occupied, W represents individual total amount.
Preferably, calculating the image information entropy of piece image by formula below：
Wherein, S is image information entropy, and N is the gray level of image, and P (i, j) is gray level co-occurrence matrixes, and i and j are represented respectively
The row coordinate and row coordinate of a pixel in image.
Preferably, calculating the image border ratio of piece image by formula below：
R=pedge/ (M × N),
Wherein, R is image border ratio, PedgeFor the number of edge pixel in image, M, N are respectively the row of image pixel
Number and columns.
Preferably, calculating the image contrast G of piece image by formula below：
Wherein, N is the gray level of image, and P (i, j) is gray level co-occurrence matrixes, and i and j represent a picture in image respectively
The row coordinate and row coordinate of element.
Preferably, calculating the image correlativity of piece image by formula below：
Wherein, COV is image correlativity, μx、μy、δxAnd δyIt is P respectivelyxAnd PyAverage and standard deviation, N be image ash
Level is spent, P (i, j) is gray level co-occurrence matrixes, and i and j represent the row coordinate and row coordinate of a pixel in image respectively,WithIt is gray level co-occurrence matrixes each column and every row pixel sum respectively.
Preferably, calculating the image energy of piece image by formula below：
Wherein, J is image energy, and N is the gray level of image, and P (i, j) is gray level co-occurrence matrixes, and i and j represent figure respectively
The row coordinate and row coordinate of a pixel as in.
Preferably, complexity statistical result is：
(1) the image complexity C1 of pure marine background image is：0<C1≤1；
(2) the image complexity C2 of the meteorological marine background image such as greasy weather, drizzle is：1<C2≤4；
(3) the image complexity C3 of the marine background image under the covering of part cloud is：C3 ＞ 4.
Preferably, the image complexity according to the image determines that the rank of the image is：
When the image complexity of image is less than or equal to 1, the image is 0 grade of image；
When the image complexity of image is more than 1 and is less than or equal to 4, the image is 1 grade of image；
When the image complexity of image is more than or equal to 4, the image is 2 grades of images.
As above it is visible, in the inventive solutions, due to the first marine background infrared image based on actual measurement, calculate sample
The characteristics of image for every piece image that notebook data is concentrated is (for example, image information entropy, image border ratio, image contrast, image phase
Guan Du and image energy), and calculate the image complexity for obtaining each width image；Then the image of the multiple image of pre-selection is answered again
Miscellaneous degree is counted, and obtains statistical result；Last image complexity and complexity statistical result further according to image determines every
The rank of piece image, so as to carry out image grading according to image complexity, and then can be in the application such as target identification
Image recognition processing algorithm is greatly simplified, computational efficiency is improved, shortens the calculating time, reduction calculates cost.
Brief description of the drawings
Fig. 1 be the embodiment of the present invention in the image grading method based on infrared image complexity flow chart.
Embodiment
For technical scheme and advantage is more clearly understood, below in conjunction with drawings and the specific embodiments, to this
Invention is described in further detail.
Fig. 1 be the embodiment of the present invention in the image grading method based on infrared image complexity flow chart.Such as Fig. 1 institutes
Show, the image grading method based on infrared image complexity in the embodiment of the present invention includes step as described below：
Step 101, the sample image sequence that need to be measured is selected, sample data set is used as.
In the inventive solutions, the sample image sequence that need to be measured can be selected first, and by the selected sample
This image sequence is used as sample data set.
In addition, preferably, in one particular embodiment of the present invention, the sample image sequence can be：Carry on the back ocean
Scape infrared image sequence.
Therefore, the more rich image of background can be included in the sample image sequence, for example, pure marine background figure
Marine background image under picture, greasy weather, the marine background image of drizzle, the covering of part cloud etc..
Step 102, the every piece image concentrated for sample data, calculates the characteristics of image of the image respectively, and according to
Obtained characteristics of image is calculated, the image complexity for obtaining image is calculated.
In the inventive solutions, the things that can be distinguished determined within the specific limits, if as an entirety
From the point of view of, it is possible to it is referred to as Generalized Sets.Wherein, each things in Generalized Sets is referred to as the element of the Generalized Sets, Generalized Set
The element of conjunction has certainty, heterogeneite, 3 features of randomness.What the complexity of one Generalized Sets was reflected is the broad sense
Gather internal element species and it is various types of included in element number how much etc. feature.
Therefore, in this step, the Generalized Sets can then be calculated using sample data set as a Generalized Sets
Image complexity in (i.e. sample data set) per piece image.
In addition, preferably, in one particular embodiment of the present invention, Generalized Sets can be calculated by formula below
Complexity (for example, image complexity of piece image)：
Wherein, C is complexity (for example, image complexity of piece image), and k is (i.e. sample data set in Generalized Sets
It is interior) number of unlike signal value (i.e. characteristics of image), n is value of statistical indicant, and a is the individual amount occupied, and W represents individual total amount.
In a Generalized Sets, the difference of personal feature is smaller, and C value is just smaller, and when feature is identical, C is
0。
In addition, in the inventive solutions, it is necessary to which the characteristics of image (i.e. value of statistical indicant) considered can include：Image is believed
Cease entropy, image border ratio, image contrast, image correlativity and image energy.
In addition, in the inventive solutions, the gray level of a width gray level image can be set as N, gray level co-occurrence matrixes
For P (i, j), wherein, i and j represent the row coordinate and row coordinate of a pixel in image respectively.Based on the gray scale symbiosis square
Battle array, can in the hope of each above-mentioned characteristics of image value.
For example, preferably, in one particular embodiment of the present invention, piece image can be calculated by formula below
Image information entropy：
Wherein, S is image information entropy, and N is the gray level of image, and P (i, j) is gray level co-occurrence matrixes, and i and j are represented respectively
The row coordinate and row coordinate of a pixel in image.
For example, preferably, in one particular embodiment of the present invention, piece image can be calculated by formula below
Image border ratio：
R=pedge/ (M × N),
Wherein, R is image border ratio, PedgeFor the number of edge pixel in image, M, N are respectively the row of image pixel
Number and columns.
For example, preferably, in one particular embodiment of the present invention, piece image can be calculated by formula below
Image contrast：
Wherein, G is image contrast, and N is the gray level of image, and P (i, j) is gray level co-occurrence matrixes, and i and j represent figure respectively
The row coordinate and row coordinate of a pixel as in.
For example, preferably, in one particular embodiment of the present invention, piece image can be calculated by formula below
Image correlativity：
Wherein, COV is image correlativity, μx、μy、δxAnd δyIt is P respectivelyxAnd PyAverage and standard deviation,
WithIt is gray level co-occurrence matrixes each column and every row pixel sum respectively.
For example, preferably, in one particular embodiment of the present invention, piece image can be calculated by formula below
Image energy：
Wherein, J is image energy.
Above-mentioned image information entropy, image border ratio, image contrast, image correlativity and image energy are substituted into above-mentioned
Formula (1) in, you can calculate and obtain the image complexity of piece image.
Step 103, Z width images are selected, the image complexity of selected Z width images is counted, complexity system is obtained
Count result.
In the inventive solutions, above-mentioned step 103 can be realized in several ways.Below will be with wherein
A kind of specific implementation exemplified by, technical scheme is introduced.
For example, preferably, in one particular embodiment of the present invention, the value of the Z can be 100, that is, have selected 100
Image of the width under DIFFERENT METEOROLOGICAL CONDITIONS；Then by actual experiment, the image complexity of 100 selected width images is carried out
Statistics, calculates image complexity average, can obtain following complexity statistical result：
(1) the image complexity C1 of pure marine background image is：0<C1≤1；
(2) the image complexity C2 of the meteorological marine background image such as greasy weather, drizzle is：1<C2≤4；
(3) the image complexity C3 of the marine background image under the covering of part cloud is：C3 ＞ 4.
Certainly, in the inventive solutions, the need for can be according to practical situations, predefine above-mentioned Z's
Value, can also be pre-selected out the Z width images under DIFFERENT METEOROLOGICAL CONDITIONS, the invention is not limited in this regard.
Step 104, the every piece image concentrated for sample data, according to the image complexity of the image and described multiple
Miscellaneous degree statistical result determines the rank of the image.
In this step, can be calculated according to the step 102 more than the obtained image complexity of each width image with
And the complexity statistical result obtained in step 103, to determine the rank for every piece image that sample data is concentrated.
In the inventive solutions, above-mentioned step 104 can be realized in several ways.Below will be with wherein
A kind of specific implementation exemplified by, technical scheme is introduced.
For example, preferably, in one particular embodiment of the present invention, it is described to be determined according to the image complexity of the image
The rank of the image can be：
When the image complexity of image is less than or equal to 1, the image is 0 grade of image；Now show the sea in the image
Foreign background is more pure；
When the image complexity of image is more than 1 and is less than or equal to 4, the image is 1 grade of image；
When the image complexity of image is more than or equal to 4, the image is 2 grades of images；Now show the sea in the image
Foreign background is complex background.
It is, of course, also possible to the need for according to actual conditions, determine that it is each that sample data is concentrated by another way
The rank of width image, the present invention is not restricted to this.
In summary, in the inventive solutions, due to the first marine background infrared image based on actual measurement, sample is calculated
The characteristics of image for every piece image that notebook data is concentrated is (for example, image information entropy, image border ratio, image contrast, image phase
Guan Du and image energy), and calculate the image complexity for obtaining each width image；Then the image of the multiple image of pre-selection is answered again
Miscellaneous degree is counted, and obtains statistical result；Last image complexity and complexity statistical result further according to image determines every
The rank of piece image, so as to carry out image grading according to image complexity, and then can be in the application such as target identification
Image recognition processing algorithm is greatly simplified, computational efficiency is improved, shortens the calculating time, reduction calculates cost.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
God is with principle, and any modification, equivalent substitution and improvements done etc. should be included within the scope of protection of the invention.
Claims (10)
1. a kind of image grading method based on infrared image complexity, it is characterised in that this method includes：
The selected sample image sequence that need to be measured, is used as sample data set；
The every piece image concentrated for sample data, calculates the characteristics of image of the image, and the figure obtained according to calculating respectively
As feature, the image complexity for obtaining image is calculated；
Z width images are selected, the image complexity of selected Z width images is counted, complexity statistical result is obtained；
The every piece image concentrated for sample data, according to the image complexity of the image and the complexity statistical result
Determine the rank of the image.
2. according to the method described in claim 1, it is characterised in that described image feature includes：
Image information entropy, image border ratio, image contrast, image correlativity and image energy.
3. method according to claim 2, it is characterised in that the image for calculating piece image by formula below is complicated
Degree：
<mrow>
<mi>C</mi>
<mo>=</mo>
<mo>-</mo>
<munderover>
<mo>&Sigma;</mo>
<mi>a</mi>
<mi>k</mi>
</munderover>
<msub>
<mi>n</mi>
<mi>a</mi>
</msub>
<mi>l</mi>
<mi>o</mi>
<mi>g</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>n</mi>
<mi>a</mi>
</msub>
<mo>/</mo>
<mi>W</mi>
<mo>)</mo>
</mrow>
<mo>;</mo>
</mrow>
Wherein, C is the image complexity of image, and k is the number of different images feature in sample data set, and n is characteristics of image, a
For the individual amount occupied, W represents individual total amount.
4. method according to claim 3, it is characterised in that the image information of piece image is calculated by formula below
Entropy：
<mrow>
<mi>S</mi>
<mo>=</mo>
<mo>-</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mi>log</mi>
<mi> </mi>
<mi>p</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
Wherein, S is image information entropy, and N is the gray level of image, and P (i, j) is gray level co-occurrence matrixes, and i and j represent image respectively
In a pixel row coordinate and row coordinate.
5. method according to claim 3, it is characterised in that the image border of piece image is calculated by formula below
Ratio：
R=pedge/ (M × N),
Wherein, R is image border ratio, PedgeFor the number of edge pixel in image, M, N be respectively image pixel line number and
Columns.
6. method according to claim 3, it is characterised in that the image contrast of piece image is calculated by formula below
G：
<mrow>
<mi>G</mi>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>-</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mi>p</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
Wherein, N is the gray level of image, and P (i, j) is gray level co-occurrence matrixes, and i and j represent a pixel in image respectively
Row coordinate and row coordinate.
7. method according to claim 3, it is characterised in that the image for calculating piece image by formula below is related
Degree：
<mrow>
<mi>C</mi>
<mi>O</mi>
<mi>V</mi>
<mo>=</mo>
<mo>&lsqb;</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<mi>i</mi>
<mi>j</mi>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msub>
<mi>&mu;</mi>
<mi>x</mi>
</msub>
<msub>
<mi>&mu;</mi>
<mi>y</mi>
</msub>
<mo>&rsqb;</mo>
<mo>/</mo>
<msub>
<mi>&delta;</mi>
<mi>x</mi>
</msub>
<msub>
<mi>&delta;</mi>
<mi>y</mi>
</msub>
<mo>,</mo>
</mrow>
Wherein, COV is image correlativity, μx、μy、δxAnd δyIt is P respectivelyxAnd PyAverage and standard deviation, N be image gray scale
Level, P (i, j) is gray level co-occurrence matrixes, and i and j represent the row coordinate and row coordinate of a pixel in image respectively,WithIt is gray level co-occurrence matrixes each column and every row pixel sum respectively.
8. method according to claim 2, it is characterised in that the image energy of piece image is calculated by formula below
Amount：
<mrow>
<mi>J</mi>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<mi>p</mi>
<msup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>,</mo>
</mrow>
Wherein, J is image energy, and N is the gray level of image, and P (i, j) is gray level co-occurrence matrixes, and i and j are represented in image respectively
A pixel row coordinate and row coordinate.
9. according to the method described in claim 1, it is characterised in that complexity statistical result is：
(1) the image complexity C1 of pure marine background image is：0<C1≤1；
(2) the image complexity C2 of the meteorological marine background image such as greasy weather, drizzle is：1<C2≤4；
(3) the image complexity C3 of the marine background image under the covering of part cloud is：C3 ＞ 4.
10. method according to claim 9, it is characterised in that the image complexity according to the image determines the figure
The rank of picture is：
When the image complexity of image is less than or equal to 1, the image is 0 grade of image；
When the image complexity of image is more than 1 and is less than or equal to 4, the image is 1 grade of image；
When the image complexity of image is more than or equal to 4, the image is 2 grades of images.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710289572.XA CN107067397A (en) | 2017-04-27 | 2017-04-27 | A kind of image grading method based on infrared image complexity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710289572.XA CN107067397A (en) | 2017-04-27 | 2017-04-27 | A kind of image grading method based on infrared image complexity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107067397A true CN107067397A (en) | 2017-08-18 |
Family
ID=59605159
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710289572.XA Pending CN107067397A (en) | 2017-04-27 | 2017-04-27 | A kind of image grading method based on infrared image complexity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107067397A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107798674A (en) * | 2017-09-15 | 2018-03-13 | 华南理工大学 | A kind of method for removing of surface of workpiece structure flase drop result |
| CN112614039A (en) * | 2020-12-25 | 2021-04-06 | 中国科学院软件研究所 | Task-driven image preprocessing algorithm evaluation method |
| CN112749620A (en) * | 2020-11-25 | 2021-05-04 | 中国电子科技集团公司第十一研究所 | Target detection method and device and readable storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615252A (en) * | 2008-06-25 | 2009-12-30 | 中国科学院自动化研究所 | A kind of method for extracting text information from adaptive images |
| CN103458772A (en) * | 2011-04-07 | 2013-12-18 | 香港中文大学 | Method and device for retinal image analysis |
| CN103824259A (en) * | 2013-11-01 | 2014-05-28 | 华南理工大学 | Image composition beautifying method and system based on visual area proportion balance rule |
| CN106097379A (en) * | 2016-07-22 | 2016-11-09 | 宁波大学 | A kind of distorted image detection using adaptive threshold and localization method |
-
2017
- 2017-04-27 CN CN201710289572.XA patent/CN107067397A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615252A (en) * | 2008-06-25 | 2009-12-30 | 中国科学院自动化研究所 | A kind of method for extracting text information from adaptive images |
| CN103458772A (en) * | 2011-04-07 | 2013-12-18 | 香港中文大学 | Method and device for retinal image analysis |
| CN103824259A (en) * | 2013-11-01 | 2014-05-28 | 华南理工大学 | Image composition beautifying method and system based on visual area proportion balance rule |
| CN106097379A (en) * | 2016-07-22 | 2016-11-09 | 宁波大学 | A kind of distorted image detection using adaptive threshold and localization method |
Non-Patent Citations (3)
| Title |
|---|
| 张学文: "组成论", 《中国科学技术大学出版社》 * |
| 张学文: "组成论介绍（中）：复杂程度和复杂度定律", 《物理与工程》 * |
| 高振宇 等: "图像复杂度描述方法研究", 《中国图象图形学报》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107798674A (en) * | 2017-09-15 | 2018-03-13 | 华南理工大学 | A kind of method for removing of surface of workpiece structure flase drop result |
| CN107798674B (en) * | 2017-09-15 | 2021-04-23 | 华南理工大学 | Method for eliminating false detection result of surface structure of metal workpiece |
| CN112749620A (en) * | 2020-11-25 | 2021-05-04 | 中国电子科技集团公司第十一研究所 | Target detection method and device and readable storage medium |
| CN112614039A (en) * | 2020-12-25 | 2021-04-06 | 中国科学院软件研究所 | Task-driven image preprocessing algorithm evaluation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106204467B (en) | Image denoising method based on cascade residual error neural network | |
| CN110048827B (en) | Class template attack method based on deep learning convolutional neural network | |
| CN111199214B (en) | Residual network multispectral image ground object classification method | |
| Jones et al. | Modelling commercial fish distributions: Prediction and assessment using different approaches | |
| CN109934826A (en) | A kind of characteristics of image dividing method based on figure convolutional network | |
| CN113496104A (en) | Rainfall forecast correction method and system based on deep learning | |
| EP3032496B1 (en) | State estimation apparatus, program, and integrated circuit | |
| CN112085056B (en) | Target detection model generation method, device, equipment and storage medium | |
| CN107067397A (en) | A kind of image grading method based on infrared image complexity | |
| CN112818969A (en) | Knowledge distillation-based face pose estimation method and system | |
| CN112013822A (en) | Multispectral remote sensing water depth inversion method based on improved GWR model | |
| CN111222583B (en) | Image steganalysis method based on countermeasure training and critical path extraction | |
| CN115359366A (en) | Remote sensing image target detection method based on parameter optimization | |
| CN115097451A (en) | Sea wave multi-parameter inversion method and system based on SAR satellite data | |
| Sivakumar | Nonlinear determinism in river flow: prediction as a possible indicator | |
| CN114821823A (en) | Image processing, training of human face anti-counterfeiting model and living body detection method and device | |
| CN104881670B (en) | A kind of fast target extracting method for SAR orientation angular estimation | |
| CN112200163B (en) | Underwater benthos detection method and system | |
| CN113989612A (en) | Remote sensing image target detection method based on attention and generation countermeasure network | |
| CN117422619A (en) | Training method of image reconstruction model, image reconstruction method, device and equipment | |
| CN116958809A (en) | Remote sensing small sample target detection method for feature library migration | |
| CN103903258B (en) | Method for detecting change of remote sensing image based on order statistic spectral clustering | |
| CN107832805B (en) | Technology for eliminating influence of spatial position error on remote sensing soft classification precision evaluation based on probability position model | |
| CN116385281A (en) | Remote sensing image denoising method based on real noise model and generated countermeasure network | |
| CN110826691A (en) | Intelligent seismic velocity spectrum pickup method based on YOLO and LSTM |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170818 |
|
| RJ01 | Rejection of invention patent application after publication |