CN105678229A - High spectral image retrieval method - Google Patents
High spectral image retrieval method Download PDFInfo
- Publication number
- CN105678229A CN105678229A CN201511020652.2A CN201511020652A CN105678229A CN 105678229 A CN105678229 A CN 105678229A CN 201511020652 A CN201511020652 A CN 201511020652A CN 105678229 A CN105678229 A CN 105678229A
- Authority
- CN
- China
- Prior art keywords
- spectrum
- curve
- point
- retrieved
- bar
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention provides a high spectral image retrieval method. The high spectral image retrieval method comprises the steps: acquiring a curve of spectrum, for a land feature, to be retrieved and a plurality of curves of spectrum requiring to be retrieved, and respectively extracting M SARI feature points at the peak and the trough of each curve of spectrum; according to the determined M SARI feature points on each curve of spectrum, acquiring N total points on each curve of spectrum through calculation; according to the N total points on each curve of spectrum, acquiring the matching point between the curve of spectrum to be retrieved and each curve of spectrum requiring to be retrieved through a band comparison method; calculating the square of the difference of the spectral values of the matching point, and according to the square of the difference of the spectral values of the matching point, calculating the similarity between the curve of spectrum to be retrieved and each curve of spectrum requiring to be retrieved; and acquiring the normalized value of the calculation result of the similarity, and outputting the retrieval results according to the sequence from large to small. The high spectral image retrieval method can effectively simplify the curve of spectrum, and can effectively improve the retrieval precision of the curve of spectrum.
Description
Technical field
The invention belongs to technical field of remote sensing image processing, particularly relate to a kind of Hyperspectral imaging search method.
Background technology
Hyperspectral imaging can provide the spectral resolution of more quantity of information and Geng Gao than common image, Objects recognition is significant, but the magnanimity meanwhile also bringing data volume increases, and how to realize video search efficiently, it it is the research focus during high-spectral data is applied. For Hyperspectral imaging, the curve of spectrum expresses the most frequently used mode of image spectral signature, therefore extracts the feature of the curve of spectrum, and adopting similarity measurement index to carry out video search is a kind of efficiently feasible method.
Curve of spectrum feature is extracted and is also widely used in the fields such as crop identification, chemical analysis. And in similarity evaluation, conventional similarity measurement index has Europe formula distance, relation conefficient, spectrum information divergence (SpectralInformationDivergence, SID), various square, spectrum angle automatching (SpectralAngleMatch, SAM) etc. But when utilizing these index of similarities to retrieve, normally all wave band datas all participate in computing, for hundreds and thousands of wave bands, calculated amount is very big, and retrieval efficiency ratio is lower.
A kind of feasible terms of settlement is that spectrum vector is carried out curve simplification. compare ripe shortcut calculation at present and have Douglas-Peucker (DP) method, hang down apart from method, Li-Openshaw method etc., wherein DP algorithm is the most conventional, its ultimate principle is: the first and last 2 of curve is virtually connected a straight line, calculated curve arrives a little the distance of straight line, and find maximum range value dmax to compare with threshold value D, if dmax is < D, then the intermediate point on this curve is all cast out, otherwise retain the point that dmax is corresponding, and taking this point as boundary, curve is divided into two portions, then these two portions are repeated operation above, until not having distance to be greater than the point of threshold value.
Although DP algorithm can preserve the morphological specificity of curve on the whole effectively, but to be used recurrence algorithm to realize, more complicated, two defects below simultaneously also having in application:
(1) needing to specify threshold value, and the selection of the effect of curve simplification and threshold value has very big relation, when threshold value increases, the point of reservation will reduce, it is possible to distortion occurs;Otherwise the point retained will increase, do not reach the object simplifying curve. In order to reach set compression effect, it is necessary to constantly attempting threshold value, workload is bigger.
(2) when many curves simplify simultaneously, for ensureing that all curves can effectively be simplified, retain enough points, need different curves is set different threshold values, workload is bigger, and the very big difference of having counted that every bar curve retains, this problem is particularly outstanding in EO-1 hyperion curve simplifies and compares.
Summary of the invention
The present invention is intended to solve that calculated amount when Hyperspectral imaging is retrieved in prior art is big, curve simplifies process needs to specify threshold value, retrieve inefficient technical problem, it is provided that a kind of retrieve calculated amount little, without the need to the Hyperspectral imaging search method specifying threshold value, retrieval efficiency is high.
Embodiments of the invention provide a kind of Hyperspectral imaging search method, comprising: obtain the curve of spectrum of atural object to be retrieved and many curves of spectrum of needs retrieval, and extract X SARI unique point respectively in the position of every bar curve of spectrum crest and trough;
According to the SARI unique point that X on every bar curve of spectrum has been determined, by calculating the N number of total point obtained on every bar curve of spectrum, the N number of total point on described every bar curve of spectrum comprises X described on this curve the SARI unique point determined;
According to the N number of total point on every bar curve of spectrum, the method compared by wave band obtains the matching point of the curve of spectrum to be retrieved with every bar curve of spectrum of needs retrieval;
Calculate the spectral value of described matching point difference square, and square calculate the similarity of the curve of spectrum to be retrieved with the every bar curve of spectrum needing retrieval according to the difference of the spectral value of described matching point;
Obtain the normalized value of the calculation result of similarity, and according to Sequential output result for retrieval from big to small;
Wherein, described SARI is spectral absorption reflection index, and described X is the first preset value, and described N is the 2nd preset value, and described N > X.
Further, X the SARI unique point determined on the every bar curve of spectrum of described basis, by calculating the concrete steps of the N number of total point obtained on every bar curve of spectrum is:
Step one, in turn straight line connect every article of curve of spectrum first point, X point in first point, X put in the 2nd point ..., X point in the X point and tail point, formation X+1 bar line segment;
Step 2, the distance arriving a little each line segment calculated on the curve of spectrum corresponding to each line segment, and compare all distance values obtained, retain that point that distance value is maximum, and form X+1 point on the corresponding curve of spectrum;
Step 3, according on the described curve of spectrum formed X+1 point, repeating step one and step 2, what retain to the curve of spectrum described in road counts as N.
Further, when chosen distance retains the most a little louder on each curve by calculating, if the distance of multiple point is equal to the distance of corresponding line segment, then retain one of them distance by following step the most a little bigger:
Obtain each point and arrive the distance value of described corresponding line segment and the length value of corresponding line segment;
Calculate and compare distance value and the worth ratio size of described homologous pair segment length that each point described arrives described corresponding line segment, and obtain the point that ratio is minimum;
Retain the point on the minimum curve of spectrum of described ratio.
Further, according to the N number of total point on every bar curve of spectrum, the step of the matching point that the method compared by wave band obtains every bar curve of spectrum that the curve of spectrum to be retrieved is retrieved with needs is:
The wave band of the curve of spectrum to be retrieved with the some place needing to retain on every bar curve of spectrum of retrieval is calculated by wave band, if the current unique point on each wave band is on same wave band, then these two Point matching.
Further, the difference of the spectral value of the described matching point of described calculating square, and the step square calculating the curve of spectrum to be retrieved and the similarity of the every bar curve of spectrum needing retrieval according to the difference of the spectral value of described matching point is:
The similarity of every bar curve of spectrum that the curve of spectrum to be retrieved is retrieved with needs
Wherein, C is the number of matching point, RA_iFor the spectral value of i-th point on the curve of spectrum to be retrieved, RB_iFor needing the spectral value retrieving i-th point on the curve of spectrum.
Further, described spectral absorption reflection indexWherein M is SARI unique point on the curve of spectrum of correspondence, ρMFor the reflectivity of M point, ρS1、ρS2Representing the left shoulder point of M point, the reflectivity of right shoulder point on the curve of spectrum respectively, d is symmetry parameter.
Further, described symmetry parameterW=λS2-λS1;
Wherein, λS1、λS2And λMRepresent the wavelength of S1, S2 and M point respectively.
Further, the concrete grammar of the curve of spectrum of described acquisition atural object to be retrieved and many curves of spectrum of needs retrieval is: directly extracts from the high spectrum image of corresponding atural object and obtains.
In above technical scheme, adopt X the SARI unique point determined on the curve of spectrum, the N number of total point obtained on every bar curve of spectrum by calculating, and according to N number of total point, the method compared by wave band obtains the matching point of the curve of spectrum to be retrieved with every bar curve of spectrum of needs retrieval, and then by the difference of the spectral value that calculates described matching point square and obtain the similarity of the curve of spectrum to be retrieved with the every bar curve of spectrum needing retrieval, on the basis of physical features accurately describing the curve of spectrum based on spectral absorption reflection index, can effectively simplify the curve of spectrum, and effectively improve the retrieval precision of the curve of spectrum.
Accompanying drawing explanation
Fig. 1 is that SARI of the present invention absorbs unique point valley point and two shoulder schematic diagram;
Fig. 2 is that SARI of the present invention absorbs unique point peak point and two shoulder schematic diagram;
Fig. 3 is the SARI unique point schematic diagram that the present invention extracts on the curve of spectrum in forest land;
Fig. 4 is the SARI unique point schematic diagram that the present invention extracts on the curve of spectrum of wheat;
Fig. 5 is the simplification Dependence Results schematic diagram in the forest land that the algorithm according to the present invention obtains;
Fig. 6 is the simplification Dependence Results schematic diagram of the wheat that the algorithm according to the present invention obtains;
Fig. 7 is the soybean of Indian remote sensing test site, the Indiana, USA northwestward, the curve of spectrum schematic diagram of wheat and forest land three kinds of atural objects of in June, 1992 shooting;
Fig. 8 is the Hyperspectral imaging search method workflow schematic diagram of an embodiment of the present invention.
Embodiment
In order to make technical problem solved by the invention, technical scheme and useful effect clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated. It is to be understood that specific embodiment described herein is only in order to explain the present invention, it is not intended to limit the present invention.
Shown in composition graphs 1 and Fig. 2, the curve of spectrum of different atural object, it is also different that it absorbs the attributes such as Wave crest and wave trough shape, position, width, the degree of depth and symmetry degree, spectral absorption reflection index SARI can be utilized accordingly to describe the recognition feature of the curve of spectrum, and it inherently have expressed the variation characteristic that object spectrum absorbs reflection coefficient.
The described curve of spectrum directly can extract from the high spectrum image of corresponding atural object and obtain, spectral absorption reflectance signature can be made up of spectral absorption two shoulder S1 and S2 on spectral absorption valley point or peak point M and the curve of spectrum, in figure, H represents the spectral absorption degree of depth, it is the distance absorbing valley point M and two shoulder composition " non-absorbing baseline ", makes ρS1、ρS2And ρMRepresent the reflectivity absorbing left shoulder S1, absorbing right shoulder point S2 and absorption point M respectively, λS1、λS2And λMRepresent left shoulder point S1 respectively, absorb the wavelength of right shoulder point S2 and absorption point M, then absorption bands width is W=λS2-λS1, symmetry parameter isAbsorbing shoulder end reflection differences is: △ ρS=ρS2-ρS1, then " non-absorbing baseline " equation is:
W·ρ-△ρSλ=ρS1-△ρS·λS1
Upper formula have expressed the spectral contribution without spectral absorption feature atural object and spectrum behavior.Therefore spectral absorption reflection index SARI may be defined as absorption or the spectral value of reflection position and the ratio of corresponding baseline value:
As shown in Figure 8, embodiments provide a kind of Hyperspectral imaging search method, comprise the steps:
Obtain the curve of spectrum of atural object to be retrieved and need many curves of spectrum of retrieval, and extract X SARI unique point respectively in the position of every bar curve of spectrum crest and trough;
Shown in composition graphs 3 and Fig. 4, Fig. 3 and Fig. 4 is respectively the curve of spectrum of forest land and wheat in AVIRIS Hyperspectral imaging, is extracted 10 SARI unique points on crest and trough respectively, and X value here is 10. In actual, the value of X artificially sets as required, the individual numerical value that the maximum value of setting can not be greater than on the curve of spectrum crest and trough.
According to the SARI unique point that X on every bar curve of spectrum has been determined, by calculating the N number of total point obtained on every bar curve of spectrum, the N number of total point on described every bar curve of spectrum comprises X described on this curve the SARI unique point determined;
According to the N number of total point on every bar curve of spectrum, the method compared by wave band obtains the matching point of the curve of spectrum to be retrieved with every bar curve of spectrum of needs retrieval;
Calculate the spectral value of described matching point difference square, and square calculate the similarity of the curve of spectrum to be retrieved with the every bar curve of spectrum needing retrieval according to the difference of the spectral value of described matching point;
Obtain the normalized value of the calculation result of similarity, and according to Sequential output result for retrieval from big to small;
Wherein, described X is the first preset value, and described N is the 2nd preset value, and described N > X.
The present invention is directed to the defect of existing DP algorithm, it is proposed that a kind of new innovatory algorithm, it is not necessary to setting threshold value, can reach and both retain enough points, can keep again the object of many curve forms simultaneously. Its basic step is as follows:
(1) connect the first point S on the curve of spectrum and end point E, calculated curve arrives a little the distance of line segment SE, and finds the some O of ultimate range, this point is retained.
(2) taking an O as boundary, connecting SO and OE, curve is divided into two portions, on calculated curve, two portions point is to the distance of corresponding line segment SO and OE, selects wherein ultimate range point P to retain.
(3) again taking O and P as boundary, curve is divided into three part SO, OP, PE, repeats the operation of (2), analogize with this, only retain one apart from maximum point every time, until the requirement counted and reach set retained.
The basis of design of the final points N retained: counting of reservation is many, calculated amount is just greatly a bit; Counting of retaining is few, and calculated amount is just a little bit smaller. But can not be excessively little, otherwise precision will be very poor, and counting of retaining so concrete needs according to practical situation, artificially sets.
(4) when chosen distance is the most a little bigger, if the distance of two points or multiple point is equal, then calculating these points to the corresponding distance of line segment and the ratio of line segment length, the point selecting ratio little is retained.
As long as it may be seen that specify the number of the point needing reservation, inventive algorithm can be restrained completely, thus ensures the effect that curve simplifies. Simultaneously owing to not having threshold value, no matter the form of curve itself how, and inventive algorithm can process on an equal basis, simplifies the morphological specificity that curve can correctly reflect curve, solves the problem of traditional algorithm when many curves simplify simultaneously. In addition in computation process, without the need to the distance of the point on double counting curve to corresponding line segment, the curve interval to last retention point place is only needed to recalculate distance, the curve in other intervals can directly calculation result according to last time obtain to the distance of corresponding line segment, therefore, can effectively reduce calculated amount.
Preferably, for embodiments of the invention, for the curve of spectrum, some point describing spectral signature needs to retain, such as some crest, trough etc., and the present invention utilizes SARI unique point to carry out the process of constraint curve simplification, first some SARI unique points on curve are extracted, using these points as initial value, in the improvement DP algorithm that the present invention proposes so that simplify the absorption feature that curve retains original spectrum curve as much as possible. Therefore, according to the SARI unique point that X on every bar curve of spectrum has been determined, by calculating the concrete steps of the N number of total point obtained on every bar curve of spectrum it is:
Step one, in turn straight line connect every article of curve of spectrum first point, X point in first point, X put in the 2nd point ..., X point in the X point and tail point, formation X+1 bar line segment;
Step 2, the distance arriving a little each line segment calculated on the curve of spectrum corresponding to each line segment, and compare all distance values obtained, retain that point that distance value is maximum, and form X+1 point on the corresponding curve of spectrum;
Step 3, according on the described curve of spectrum formed X+1 point, repeating step one and step 2, what retain to the curve of spectrum described in road counts as N.
Shown in composition graphs 3, the curve of spectrum in Fig. 3 is extracted 20 SARI unique points, comprise 10 SARI absorption peak points and 10 SARI absorption valley points, the method utilizing the present invention described above simplifies, result simplification curve as shown in Figures 5 and 6, always counting that wherein Fig. 5 retains is 50, be exactly also described N is set to 50; Fig. 6 retains always to count be 100, namely N value here is 100, compares with Fig. 3 afterwards it may be seen that the inventive method can intactly retain the configuration feature of curve.
Preferably, in above-mentioned steps, when chosen distance retains the most a little louder on each curve by calculating, if the distance of multiple point is equal to the distance of corresponding line segment, then retain one of them distance by following step the most a little bigger:
Obtain each point and arrive the distance value of described corresponding line segment and the length value of corresponding line segment;
Calculate and compare distance value and the worth ratio size of described homologous pair segment length that each point described arrives described corresponding line segment, and obtain the point that ratio is minimum;
Retain the point on the minimum curve of spectrum of described ratio.
Further, according to the N number of total point on every bar curve of spectrum, the step of the matching point that the method compared by wave band obtains every bar curve of spectrum that the curve of spectrum to be retrieved is retrieved with needs is:
The wave band of the curve of spectrum to be retrieved with the some place needing to retain on every bar curve of spectrum of retrieval is calculated by wave band, if the current unique point on each wave band is on same wave band, then these two Point matching.
Further, the difference of the spectral value of the described matching point of described calculating square, and the step square calculating the curve of spectrum to be retrieved and the similarity of the every bar curve of spectrum needing retrieval according to the difference of the spectral value of described matching point is:
The similarity of every bar curve of spectrum that the curve of spectrum to be retrieved is retrieved with needs
Wherein, C is the number of matching point, RA_iFor the spectral value of i-th point on the curve of spectrum to be retrieved, RB_iFor needing the spectral value retrieving i-th point on the curve of spectrum.
Every bar of above-mentioned calculating needs the curve of spectrum of retrieval and the similarity of the curve of spectrum to be retrieved are normalized, and according to Sequential output result for retrieval from big to small, retrieval can be completed.
In order to verify the validity of search method of the present invention, as shown in Figure 7, test with the AVIRIS Hyperspectral imaging in Purdue university remote sensing image processing laboratory, a part for the Indian remote sensing test site, the Indiana, USA northwestward of in June, 1992 shooting taken from by original image, there are 224 wave bands, spectral range is 0.4 to 2.45 μm, spatial resolution is 20 meters, image size is 145*145 pixel, after removing water absorption bands and bigger wave band affected by noise, obtain 183 wave bands as research object.
With reference to true line map, provide the result for retrieval that three classes easily mix atural object (soybean, wheat and forest land), three of this three classes every class of atural object are analyzed as unit, simultaneously SID and SAM two kinds of methods with routine compare, result is as shown in Table 1 to Table 3, wherein A1, A2, A3 represent soybean, and B1, B2, B3 and C1, C2, C3 represent wheat and forest land respectively, and this paper method remains 50 on the curve of spectrum point.
Wherein, oblique runic represents the result of retrieval mistake.
From above-mentioned Experimental comparison it may be seen that the method for the present invention can be entirely true distinguish different atural object, and there is a lot of mistake in two kinds of now general methods, this also illustrates the validity of search method of the present invention.
Utilize the technical scheme of the present invention, the physical features of the curve of spectrum can be described more accurately based on spectral absorption reflection index, the unique point of identical number can be obtained adaptively based on improvement DP algorithm, adopt compared with whole wave band carries out retrieval with tradition search method, the present invention can realize the retrieval of high precision by simplifying curve.
The foregoing is only the better embodiment of the present invention, not in order to limit the present invention, all any amendment, equivalent replacement and improvement etc. done within the spirit and principles in the present invention, all should be included within protection scope of the present invention.
Claims (8)
1. a Hyperspectral imaging search method, it is characterised in that: comprise,
Obtain the curve of spectrum of atural object to be retrieved and need many curves of spectrum of retrieval, and extract X SARI unique point respectively in the position of every bar curve of spectrum crest and trough;
According to the SARI unique point that X on every bar curve of spectrum has been determined, by calculating the N number of total point obtained on every bar curve of spectrum, the N number of total point on described every bar curve of spectrum comprises X described on this curve the SARI unique point determined;
According to the N number of total point on every bar curve of spectrum, the method compared by wave band obtains the matching point of the curve of spectrum to be retrieved with every bar curve of spectrum of needs retrieval;
Calculate the spectral value of described matching point difference square, and square calculate the similarity of the curve of spectrum to be retrieved with the every bar curve of spectrum needing retrieval according to the difference of the spectral value of described matching point;
Obtain the normalized value of the calculation result of similarity, and according to Sequential output result for retrieval from big to small;
Wherein, described SARI is spectral absorption reflection index, and described X is the first preset value, and described N is the 2nd preset value, and described N > X.
2. Hyperspectral imaging search method according to claim 1, it is characterised in that: X the SARI unique point determined on the every bar curve of spectrum of described basis, by calculating the concrete steps of the N number of total point obtained on every bar curve of spectrum is:
Step one, in turn straight line connect every article of curve of spectrum first point, X point in first point, X put in the 2nd point ..., X point in the X point and tail point, formation X+1 bar line segment;
Step 2, the distance arriving a little each line segment calculated on the curve of spectrum corresponding to each line segment, and compare all distance values obtained, retain that point that distance value is maximum, and form M+1 point on the corresponding curve of spectrum;
Step 3, according on the described curve of spectrum formed X+1 point, repeating step one and step 2, what retain to the curve of spectrum described in road counts as N.
3. Hyperspectral imaging search method according to claim 2, it is characterized in that: when by calculating, on each curve, chosen distance retains the most a little louder, if the distance of multiple point is equal to the distance of corresponding line segment, then retain one of them distance by following step the most a little bigger:
Obtain each point and arrive the distance value of described corresponding line segment and the length value of corresponding line segment;
Calculate and compare distance value and the worth ratio size of described homologous pair segment length that each point described arrives described corresponding line segment, and obtain the point that ratio is minimum;
Retain the point on the minimum curve of spectrum of described ratio.
4. Hyperspectral imaging search method according to claim 1, it is characterised in that: according to the N number of total point on every bar curve of spectrum, the step of the matching point that the method compared by wave band obtains every bar curve of spectrum that the curve of spectrum to be retrieved is retrieved with needs is:
The wave band of the curve of spectrum to be retrieved with the some place needing to retain on every bar curve of spectrum of retrieval is calculated by wave band, if the current unique point on each wave band is on same wave band, then these two Point matching.
5. Hyperspectral imaging search method according to claim 1, it is characterized in that: the difference of the spectral value of the described matching point of described calculating square, and the step square calculating the curve of spectrum to be retrieved and the similarity of the every bar curve of spectrum needing retrieval according to the difference of the spectral value of described matching point is:
The similarity of every bar curve of spectrum that the curve of spectrum to be retrieved is retrieved with needs
Wherein, C is the number of matching point, RA_iFor the spectral value of i-th point on the curve of spectrum to be retrieved, RB_iFor needing the spectral value retrieving i-th point on the curve of spectrum.
6. Hyperspectral imaging search method according to claim 1, it is characterised in that: described spectral absorption reflection indexWherein M is SARI unique point on the curve of spectrum of correspondence, ρMFor the reflectivity of M point, ρS1、ρS2Representing the left shoulder point of M point, the reflectivity of right shoulder point on the curve of spectrum respectively, d is symmetry parameter.
7. Hyperspectral imaging search method according to claim 6, it is characterised in that: described symmetry parameterW=λS2-λS1;
Wherein, λS1、λS2And λMRepresent the wavelength of S1, S2 and M point respectively.
8. Hyperspectral imaging search method according to claim 1, it is characterised in that: the concrete grammar of the curve of spectrum of described acquisition atural object to be retrieved and many curves of spectrum of needs retrieval is: directly extracts from the high spectrum image of corresponding atural object and obtains.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511020652.2A CN105678229B (en) | 2015-12-29 | 2015-12-29 | A kind of Hyperspectral imaging search method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511020652.2A CN105678229B (en) | 2015-12-29 | 2015-12-29 | A kind of Hyperspectral imaging search method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105678229A true CN105678229A (en) | 2016-06-15 |
| CN105678229B CN105678229B (en) | 2019-03-05 |
Family
ID=56298083
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201511020652.2A Active CN105678229B (en) | 2015-12-29 | 2015-12-29 | A kind of Hyperspectral imaging search method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105678229B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110974191A (en) * | 2019-12-18 | 2020-04-10 | 深圳市精气生物技术有限公司 | Radial artery waveform data identification algorithm |
| CN111324805A (en) * | 2018-12-13 | 2020-06-23 | 北京搜狗科技发展有限公司 | Query intention determining method and device, searching method and searching engine |
| CN113792082A (en) * | 2021-09-02 | 2021-12-14 | 深圳创景数科信息技术有限公司 | Fabric component retrieval method based on database |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101916377A (en) * | 2010-07-15 | 2010-12-15 | 北京航空航天大学 | Method for intelligently identifying mineral information based on spectral information |
| CN102609711A (en) * | 2012-02-21 | 2012-07-25 | 核工业北京地质研究院 | Information extraction method applicable to hyperspectral image |
| CN104331482A (en) * | 2014-11-07 | 2015-02-04 | 武汉大学 | Method and system for hyper-spectrum matching based on feature index |
-
2015
- 2015-12-29 CN CN201511020652.2A patent/CN105678229B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101916377A (en) * | 2010-07-15 | 2010-12-15 | 北京航空航天大学 | Method for intelligently identifying mineral information based on spectral information |
| CN102609711A (en) * | 2012-02-21 | 2012-07-25 | 核工业北京地质研究院 | Information extraction method applicable to hyperspectral image |
| CN104331482A (en) * | 2014-11-07 | 2015-02-04 | 武汉大学 | Method and system for hyper-spectrum matching based on feature index |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111324805A (en) * | 2018-12-13 | 2020-06-23 | 北京搜狗科技发展有限公司 | Query intention determining method and device, searching method and searching engine |
| CN111324805B (en) * | 2018-12-13 | 2024-02-13 | 北京搜狗科技发展有限公司 | Query intention determining method and device, searching method and searching engine |
| CN110974191A (en) * | 2019-12-18 | 2020-04-10 | 深圳市精气生物技术有限公司 | Radial artery waveform data identification algorithm |
| CN113792082A (en) * | 2021-09-02 | 2021-12-14 | 深圳创景数科信息技术有限公司 | Fabric component retrieval method based on database |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105678229B (en) | 2019-03-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109543630B (en) | Remote sensing image woodland extraction method and system based on deep learning, storage medium and electronic equipment | |
| CN108446312B (en) | Optical remote sensing image retrieval method based on deep convolution semantic net | |
| CN104268883A (en) | Time-frequency spectrum curve extracting method based on edge detection | |
| CN105809140A (en) | Method and device for extracting surface water body information based on remote sensing model | |
| CN107832849B (en) | Knowledge base-based power line corridor three-dimensional information extraction method and device | |
| CN105224961B (en) | A kind of infrared spectrum feature extracting and matching method of high resolution | |
| CN103714341B (en) | Hyper spectral reflectance data light spectrum signature extracting method based on overall situation sensitivity analysis | |
| CN105528580B (en) | A kind of EO-1 hyperion Curve Matching method based on absorption peak feature | |
| US20170154056A1 (en) | Matching image searching method, image searching method and devices | |
| CN105678229A (en) | High spectral image retrieval method | |
| CN103279636B (en) | A kind of method determining tree age | |
| CN102298705A (en) | Method for analyzing influence of landscape characteristic on remote sensing classification patch accuracy | |
| CN105574265B (en) | Entire assembly model quantitative description towards model index | |
| CN110263735A (en) | A method of tree species classification being carried out to artificial forest high-spectral data using Three dimensional convolution neural network | |
| CN106198485A (en) | A kind of Raman spectrum matching process | |
| CN103745453A (en) | Town information extraction method based on Google Earth remote sensing image | |
| CN111709775A (en) | House property price evaluation method and device, electronic equipment and storage medium | |
| CN103868858B (en) | A kind ofly determine the method that saliferous clay dominates salinity spectral response best band | |
| CN106935038A (en) | One kind parking detecting system and detection method | |
| CN103196356A (en) | Oil tube defect quantitative recognition method based on support vector machine | |
| CN104361192A (en) | Sparse representation adaptive reconstruction method under compressed sensing analysis model | |
| CN104484679A (en) | Non-standard gun shooting bullet trace image automatic identification method | |
| Ghiyamat et al. | Airborne hyperspectral discrimination of tree species with different ages using discrete wavelet transform | |
| CN104715160A (en) | Soft measurement modeling data outlier detecting method based on KMDB | |
| CN106960119A (en) | A kind of fusion DI forest disturbance comprehensive characteristics index and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |