CN110533671A - A kind of method of local dynamic station planning cutting literal line - Google Patents

A kind of method of local dynamic station planning cutting literal line Download PDF

Info

Publication number
CN110533671A
CN110533671A CN201910763994.5A CN201910763994A CN110533671A CN 110533671 A CN110533671 A CN 110533671A CN 201910763994 A CN201910763994 A CN 201910763994A CN 110533671 A CN110533671 A CN 110533671A
Authority
CN
China
Prior art keywords
unit
text
scheme
width
question open
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
Application number
CN201910763994.5A
Other languages
Chinese (zh)
Inventor
罗胜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wenzhou University
Original Assignee
Wenzhou University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Wenzhou University filed Critical Wenzhou University
Priority to CN201910763994.5A priority Critical patent/CN110533671A/en
Publication of CN110533671A publication Critical patent/CN110533671A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • 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
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • 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
    • G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
    • 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
    • G06T2207/20—Special algorithmic details
    • G06T2207/20112—Image segmentation details
    • G06T2207/20156—Automatic seed setting

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Machine Translation (AREA)

Abstract

The invention discloses a kind of methods that local dynamic station plans cutting literal line, and the unit that will continuously leave a question open is as region of leaving a question open, with the L unit U that leave a question openi, including the previous detection text in region of leaving a question open and region the latter detection text that leaves a question open, total L+2 unit constitute and leave a question open unit collection U；The matrix that (L+2) × (L+2) is constructed with this L+2 unit, calculates point (Uh,Ue) constitute character at word value at cost Phe, after in matrix at word cost normalized, upper right triangular matrix width be 4 belt-like zone in carry out Dynamic Programming, find optimal case.The present invention only handles continuous suspicious unit, only attempts combining unit by the place of over-segmentation there may be characters, therefore be the global analysis with multiple local analyses instead of entire row, big memory is not needed, it economizes on resources, speed is fast, and effect is more preferable.

Description

A kind of method of local dynamic station planning cutting literal line
Technical field
The present invention relates to technical field of character recognition, and in particular to a kind of method of local dynamic station planning cutting literal line.
Background technique
For a line being cut into image or a column text, it is desirable to which the text in home row generally uses sciagraphy. Such as a line text shown in Fig. 1 (a), add up all rows along longitudinal direction, and projection becomes one-dimensional data, as shown in Fig. 1 (b).One In dimension data, often array sequence is cut into minimum value the unit of separation, by each unit as a character.
But in fact, Chinese character has the structures such as tiled configuration, left, center, right.In this simple sciagraphy, one or so knot Structure Chinese character may be divided into two units, for example, " print ", " ", " row ", " warp ", " item ", the words such as " remittance "；Double quotation marks ' " ' Two ' " single quotation marks may be processed into；The Chinese character of left, center, right structure can possibly even be divided into three units, such as " doing ", The words such as " tree ", " lake ", " spreading out ".If with cell spacing from judging whether multiple units belong to the same Chinese character, distance selection Must be too small, then " two " of vertical setting of types, " three " are easy to be divided into two or three " one " words, on the right of " doing " word ", " be easy to cut At punctuation mark.If distance threshold selects too big, " 11 " are also easy to be processed into a word, and the number for scheming medium and small No.1 is held It is just easily merged a word, fullstop in figure "." be also easy to be integrated into previous word.Therefore simple threshold value is relied on, threshold value is very Hardly possible is determining, and not fully accurate.
Also have and carry out combining unit in the method for Dynamic Programming, to all unit permutation and combination in row at a big matrix, In Seek optimal case in matrix.But in fact, row in some units it is sufficiently wide, word can be directly determined into, do not need with it is other Unit reattempts Combinatorial Optimization.Therefore this method wastes some significant clues, needs to occupy big memory to store, operate Big matrix, waste of resource, speed is slow, and effect is not best.
Summary of the invention
To solve the above problems, may deposited the present invention provides a kind of method that local dynamic station plans cutting literal line Combining unit is attempted by the place of over-segmentation in character, big memory is not needed, economizes on resources, speed is fast, and effect is more It is good.
To achieve the above object, the technical scheme adopted by the invention is as follows:
A kind of method of local dynamic station planning cutting literal line, intersymbol combination is searched based on the method for Dynamic Programming Optimal case, include the following steps:
S1, two points of threshold values that row/column image is calculated using Ostu method (difference method between maximum kind), are changed into two for image It is worth (black and white) image, white is the word in prospect, and black is background；
S2, using length-width ratio, foreground area of the size in text possible range as candidate text, statistics text is average Width Wc；
S3, by row text, each row projection that adds up along longitudinal direction becomes one-dimensional data；If it is column text, transversely add up each Column projection also becomes one-dimensional data；
S4, the data that one-dimensional data left and right ends are 0 are excluded, finds the minimum for having the projection of character portion among data Value；
S5, after finding the data that one-dimensional data exclusion left and right ends are 0, there are all minimum values of the projection of character portion；
S6, lowered zones are set by the region where each minimum value, finds the right boundary of lowered zones, low-lying district Region between domain is peak region, judges a possibility that lowered zones are interword gap, peak region is text unit, it would be possible to Property be more than empirical value peak region deposit text unit array, lowered zones be stored in interword gap array, it would be possible to property is low Left and right lowered zones are merged into the peak region of empirical value；
It is average to will be greater than text by S7, the text mean breadth for counting the width of each text unit divided by step S2 The text unit and interword gap of width presupposition multiple are greater than the unit of maximum text width directly as the text detected, Using other units as the unit that leaves a question open, and using continuous multiple, intermediate nothing leave a question open unit detection text as character area, calculating The average word width W of each character areacWith average interword gap Wb；
S8, the unit that will continuously leave a question open are as region of leaving a question open, with the L unit U that leave a question openi, including the previous detection in region of leaving a question open Text and region the latter detection text that leaves a question open, total L+2 unit constitute the unit collection U that leaves a question open；With this L+2 unit construction one The matrix of a (L+2) × (L+2), the point (U in matrixh,Ue)(Uh≤ Ue, e-h≤4) and it indicates from unit UhThe left side starts, In Unit UeThe right constitutes a character, point (U in the range of terminatingh,Ue) value PheIndicate this range constitute character at word Cost；
Phe=λ1(Whe-Wc)/(Whe+Wc)+λ2(Whb-Wb)/(Whb+Wb)+λ3(Web-Wb)/(Whe+Wb)；
In formula, λ1-λ3It is weighting coefficient, WheIt is unit Uh, unit UeBetween width, i.e., from unit UhLeft margin is to unit Ue The right edge distance,WhbIt is unit UhThe width in left side gap,WebIt is unit UeThe width in the right gap；This formula illustrates structure At character and left and right character similarity degree；
S9, by matrix at word cost normalized, i.e., divided by matrix at (obtaining into after word cost maximum value For character or a possibility that be merged into character), upper right triangular matrix width be 4 belt-like zone in carry out dynamic rule It draws,
Find optimal case；Optimal case is averaged into word cost minimization, and the variance of character width, interword gap width Variance it is also minimum, such as following formula:
Cost=λ4mean(Phe)+λ5δWt+λ6δWb；
In formula, λ4-λ6It is weighting coefficient, mean (Phe) it is the average at word cost, δ of all the points in schemeWtInstitute in scheme There are the variance of character width, δWbIt is the variance of all interword gap width in scheme.
Further, in the step S6, the average height of peak region should be higher than the average height of lowered zones Htmin。
Further, the step of Dynamic Programming is as follows:
(1) seed is generated: with 4 points of the first row for 4 seeds, as 4 kinds of schemes；
(2) scheme is grown: every kind of scheme is grown downwards, from point (Uh,Ue) downwards growth when, select Ue+14 capable points add Enter scheme；N kind scheme, every kind of scheme has 4 kinds may select when growing downwards, therefore grows a n kind scheme and become the kind side 4n Case；
(3) scheme is cut: being calculated the cost of 4n kind scheme, is selected the smallest m kind scheme of cost as seed scheme；
The number at scheme midpoint starts to cut for the first time when being more than 3, can improve the accuracy of algorithm；
(4) step (2), (3) are repeated, until each scheme reaches the last one unit in the unit collection U that leaves a question open；
(5) selecting the smallest scheme of cost is optimal case.
Further, the optimal case be based on known left and right character width, interword gap judgement leave a question open unit whether It should merge, merge character, the text of Embedded step S7 detection, as final result by the strategy that optimal case provides.
The invention has the following advantages:
The present invention only handles continuous suspicious unit, only there may be characters by the place of over-segmentation attempt to merge it is single Member, therefore be the global analysis with multiple local analyses instead of entire row, big memory is not needed, is economized on resources, speed Fastly, and effect is more preferable.
Detailed description of the invention
Fig. 1 is the projection of line of text and its horizontal direction in image.
Fig. 2 is into word Cost matrix.
Fig. 3 is at the dynamic programming process in word Cost matrix.
Specific embodiment
The present invention is described in detail combined with specific embodiments below.Following embodiment will be helpful to the technology of this field Personnel further understand the present invention, but the invention is not limited in any way.It should be pointed out that the ordinary skill of this field For personnel, without departing from the inventive concept of the premise, various modifications and improvements can be made.These belong to the present invention Protection scope.
The embodiment of the invention provides a kind of methods that local dynamic station plans cutting literal line, include the following steps:
S1, two points of threshold values that row/column image is calculated using Ostu method (difference method between maximum kind), are changed into two for image It is worth (black and white) image, white is the word in prospect, and black is background；
S2, using length-width ratio, foreground area of the size in text possible range as candidate text, statistics text is average Width Wc；
S3, by row text, each row projection that adds up along longitudinal direction becomes one-dimensional data；If it is column text, transversely add up each Column projection also becomes one-dimensional data；
S4, the data that one-dimensional data left and right ends are 0 are excluded, finds the minimum for having the projection of character portion among data Value；
S5, after finding the data that one-dimensional data exclusion left and right ends are 0, there are all minimum values of the projection of character portion；
S6, lowered zones are set by the region where each minimum value, finds the right boundary of lowered zones, low-lying district Region between domain is peak region, judges a possibility that lowered zones are interword gap, peak region is text unit, it would be possible to Property be more than empirical value peak region deposit text unit array, lowered zones be stored in interword gap array, it would be possible to property is low Left and right lowered zones are merged into the peak region of empirical value；Wherein, the average height of peak region should compare lowered zones The high H of average heighttmin；
It is average to will be greater than text by S7, the text mean breadth for counting the width of each text unit divided by step S2 The text unit and interword gap of width presupposition multiple are greater than the unit of maximum text width directly as the text detected, Using other units as the unit that leaves a question open, and using continuous multiple, intermediate nothing leave a question open unit detection text as character area, calculating The average word width W of each character areacWith average interword gap Wb；
S8, the unit that will continuously leave a question open are as region of leaving a question open, with the L unit U that leave a question openi, including the previous detection in region of leaving a question open Text and region the latter detection text that leaves a question open, total L+2 unit constitute the unit collection U that leaves a question open；With this L+2 unit construction one The matrix of a (L+2) × (L+2), the point (U in matrixh,Ue)(Uh≤ Ue, e-h≤4) and it indicates from unit UhThe left side starts, In Unit UeThe right constitutes a character, point (U in the range of terminatingh,Ue) value PheIndicate this range constitute character at word Cost；
Phe=λ1(Whe-Wc)/(Whe+Wc)+λ2(Whb-Wb)/(Whb+Wb)+λ3(Web-Wb)/(Whe+Wb)
In formula, λ1-λ3It is weighting coefficient, WheIt is unit Uh, unit UeBetween width, i.e., from unit UhLeft margin is to unit Ue The right edge distance,WhbIt is unit UhThe width in left side gap,WebIt is unit UeThe width in the right gap；This formula illustrates structure At character and left and right character similarity degree；
S9, by matrix at word cost normalized (divided by matrix at word cost maximum value) after, such as Fig. 3 In plus the point of Δ illustrate U4、U5A possibility that being merged into a character；If Uh=Ue, indicate UhUnit can not be with right cell Merge, individually becomes a character；
Since row is character start unit, column are character ends units, and therefore, this matrix only has upper right triangular portions；And And due to being at most divided into four units in a character horizontal direction, e-h≤4, on this upper right triangular matrix is also diagonal The belt-like zone that an only width is 4；
Dynamic Programming is carried out in the belt-like zone that the width of upper right triangular matrix is 4, finds optimal case；Optimal case It is average at word cost minimization, and the variance of the variance of character width, interword gap width is also minimum, such as following formula:
Cost=λ4mean(Phe)+λ5δWt+λ6δWb；
In formula, λ4-λ6It is weighting coefficient, mean (Phe) it is the average at word cost, δ of all the points in schemeWtInstitute in scheme There are the variance of character width, δWbIt is the variance of all interword gap width in scheme.
The step of Dynamic Programming, is as follows:
(1) seed is generated: with 4 points of the first row for 4 seeds, as 4 kinds of schemes；
(2) scheme is grown: every kind of scheme is grown downwards, such as from point (Uh,Ue) downwards growth when, select Ue+1Capable 4 Point addition scheme；N kind scheme, every kind of scheme has 4 kinds may select when growing downwards, therefore grows a n kind scheme and become 4n Kind scheme；
(3) scheme is cut: being calculated the cost of 4n kind scheme, is selected the smallest m kind scheme of cost as seed scheme；
The number at scheme midpoint starts to cut for the first time when being more than 3, can improve the accuracy of algorithm；
(4) step (2), (3) are repeated, until each scheme reaches the last one unit in the unit collection U that leaves a question open；
(5) selecting the smallest scheme of cost is optimal case；The optimal case is based on known left and right character width, word Between the gap judgement unit that leaves a question open whether should merge, such as have U in optimal case2U4Point just merges Unit 2,3,4.By optimal side The strategy that case provides merges character, the text of Embedded step S7 detection, as final result.
Specific embodiments of the present invention are described above.It is to be appreciated that the invention is not limited to above-mentioned Particular implementation, those skilled in the art can make a variety of changes or modify within the scope of the claims, this not shadow Ring substantive content of the invention.In the absence of conflict, the feature in embodiments herein and embodiment can any phase Mutually combination.

Claims (4)

1. a kind of method of local dynamic station planning cutting literal line, characterized by the following steps:
Image is changed into bianry image by S1, two points of threshold values that row/column image is calculated using Ostu method, and white is in prospect Word, black is background；
S2, using length-width ratio, foreground area of the size in text possible range as candidate text, count text mean breadth Wc；
S3, by row text, each row projection that adds up along longitudinal direction becomes one-dimensional data；If it is column text, each column that transversely add up are thrown Shadow also becomes one-dimensional data；
S4, the data that one-dimensional data left and right ends are 0 are excluded, finds the minimum value for having the projection of character portion among data；
S5, after finding the data that one-dimensional data exclusion left and right ends are 0, there are all minimum values of the projection of character portion；
S6, lowered zones are set by the region where each minimum value, finds the right boundary of lowered zones, between lowered zones Region be peak region, judge a possibility that lowered zones are interword gap, peak region is text unit, it would be possible to which property is super Cross peak region deposit text unit array, the lowered zones deposit interword gap array of empirical value, it would be possible to which property is lower than warp The peak region for testing threshold value is merged into left and right lowered zones；
S7, the text mean breadth for counting the width of each text unit divided by step S2, will be greater than text mean breadth The text unit and interword gap of presupposition multiple are greater than the unit of maximum text width directly as the text detected, by it Its unit as leaving a question open unit, and using continuous multiple, intermediate nothing leave a question open unit detection text as character area, calculate each The average word width W of character areacWith average interword gap Wb；
S8, the unit that will continuously leave a question open are as region of leaving a question open, with the L unit U that leave a question openi, including the previous detection text in region of leaving a question open and The region the latter that leaves a question open detects text, and total L+2 unit constitutes the unit collection U that leaves a question open；One (L+2) is constructed with this L+2 unit The matrix of × (L+2), the point (U in matrixh,Ue)(Uh≤ Ue, e-h≤4) and it indicates from unit UhThe left side starts, in unit UeIt is right While constituting a character, point (U in the range of terminatingh,Ue) value PheIndicate this range constitute character at word cost；
Phe=λ1(Whe-Wc)/(Whe+Wc)+λ2(Whb-Wb)/(Whb+Wb)+λ3(Web-Wb)/(Whe+Wb)；
In formula, λ1-λ3It is weighting coefficient, WheIt is unit Uh, unit UeBetween width, i.e., from unit UhLeft margin is to unit UeThe right side The distance at edge, WhbIt is unit UhThe width in left side gap, WebIt is unit UeThe width in the right gap；
S9, by matrix at word cost normalized, i.e., divided by matrix at word cost maximum value after, in upper right triangle Dynamic Programming is carried out in the belt-like zone that the width of matrix is 4, finds optimal case；Optimal case is averaged into word cost most It is small, and the variance of the variance of character width, interword gap width is also minimum, such as following formula:
Cost=λ4mean(Phe)+λ5δWt+λ6δWb；
In formula, λ4-λ6It is weighting coefficient, mean (Phe) it is the average at word cost, δ of all the points in schemeWtAll words in scheme Accord with the variance of width, δWbIt is the variance of all interword gap width in scheme.
2. a kind of method of local dynamic station planning cutting literal line as described in claim 1, it is characterised in that: the step S6 In, the average height of peak region should be H higher than the average height of lowered zonestmin。
3. a kind of method of local dynamic station planning cutting literal line as described in claim 1, it is characterised in that: the dynamic rule The step of drawing is as follows:
(1) seed is generated: with 4 points of the first row for 4 seeds, as 4 kinds of schemes；
(2) scheme is grown: every kind of scheme is grown downwards, from point (Uh,Ue) downwards growth when, select Ue+14 capable point addition sides Case；N kind scheme, every kind of scheme has 4 kinds may select when growing downwards, therefore grows a n kind scheme and become 4n kind scheme；
(3) scheme is cut: being calculated the cost of 4n kind scheme, is selected the smallest m kind scheme of cost as seed scheme；Scheme midpoint Number start when being more than 3 to cut for the first time；
(4) step (2), (3) are repeated, until each scheme reaches the last one unit in the unit collection U that leaves a question open；
(5) selecting the smallest scheme of cost is optimal case.
4. a kind of method of local dynamic station planning cutting literal line as described in claim 1, it is characterised in that: the optimal side Case is based on known left and right character width, interword gap judgement leaves a question open, and whether unit should merge, the plan provided by optimal case Slightly merge character, the text of Embedded step S7 detection, as final result.
CN201910763994.5A 2019-08-19 2019-08-19 A kind of method of local dynamic station planning cutting literal line Pending CN110533671A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910763994.5A CN110533671A (en) 2019-08-19 2019-08-19 A kind of method of local dynamic station planning cutting literal line

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910763994.5A CN110533671A (en) 2019-08-19 2019-08-19 A kind of method of local dynamic station planning cutting literal line

Publications (1)

Publication Number Publication Date
CN110533671A true CN110533671A (en) 2019-12-03

Family

ID=68663818

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910763994.5A Pending CN110533671A (en) 2019-08-19 2019-08-19 A kind of method of local dynamic station planning cutting literal line

Country Status (1)

Country Link
CN (1) CN110533671A (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011146028A (en) * 2010-01-18 2011-07-28 Fujitsu Ltd Character recognition method and character recognition device
CN104143093A (en) * 2013-05-10 2014-11-12 百度在线网络技术（北京）有限公司 Character identification method and device

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011146028A (en) * 2010-01-18 2011-07-28 Fujitsu Ltd Character recognition method and character recognition device
CN104143093A (en) * 2013-05-10 2014-11-12 百度在线网络技术（北京）有限公司 Character identification method and device

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
姚正斌, 丁晓青, 刘长松: "基于笔划合并和动态规划的联机汉字切分算法", 清华大学学报(自然科学版), no. 10, 30 October 2004 (2004-10-30), pages 1417 - 1421 *
杨晓娟;宋凯;: "基于投影法的文档图像分割算法", 成都大学学报(自然科学版), no. 02, 30 June 2009 (2009-06-30), pages 139 - 141 *

Similar Documents

Publication Publication Date Title
CN102663382B (en) Video image character recognition method based on submesh characteristic adaptive weighting
Keshtkar et al. Land-cover classification and analysis of change using machine-learning classifiers and multi-temporal remote sensing imagery
Lindsay et al. Removal of artifact depressions from digital elevation models: towards a minimum impact approach
CN106548141B (en) A kind of object-oriented farmland information extraction method based on the triangulation network
CN102509383B (en) Feature detection and template matching-based mixed number identification method
CN101515325B (en) Character extracting method in digital video based on character segmentation and color cluster
CN104504717B (en) A kind of image information detecting method and device
CN102096816B (en) Multi-scale multi-level image segmentation method based on minimum spanning tree
CN101408941A (en) Method for multi-dimension segmentation of remote sensing image and representation of segmentation result hierarchical structure
CN106407883A (en) Complex table and method for identifying handwritten numbers in complex table
CN107092871B (en) Remote sensing image building detection method based on multiple dimensioned multiple features fusion
CN103218833B (en) The color space the most steady extremal region detection method of Edge Enhancement type
CN102881017B (en) Cell separation method
CN107833224A (en) A kind of image partition method based on multi-level region synthesis
CN110533671A (en) A kind of method of local dynamic station planning cutting literal line
CN101702236B (en) Multi-target foreground segmentation method
CN107610137A (en) A kind of high-resolution remote sensing image optimal cut part method
CN110532537A (en) A method of text is cut based on two points of threshold methods and sciagraphy multistage
CN116415318B (en) Modeling method for hydrologic connectivity of lake in inner flow area based on mathematical morphology
CN104239551B (en) Multi-feature VP-tree index-based remote sensing image retrieval method and multi-feature VP-tree index-based remote sensing image retrieval device
CN110427341A (en) A kind of knowledge mapping entity relationship method for digging based on paths ordering
CN105894501A (en) Single-tree detection and crown describing method for high-resolution remote sensing image
CN109460763A (en) A kind of text area extraction method positioned based on multi-level document component with growth
CN108966241A (en) A kind of optimization method of adaptive impovement fish-swarm algorithm
CN107103611A (en) A kind of image partition method based on maximum stable homogeneous area

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