CN108509926B - Building extraction method based on bidirectional color space transformation - Google Patents

Building extraction method based on bidirectional color space transformation Download PDF

Info

Publication number
CN108509926B
CN108509926B CN201810307834.5A CN201810307834A CN108509926B CN 108509926 B CN108509926 B CN 108509926B CN 201810307834 A CN201810307834 A CN 201810307834A CN 108509926 B CN108509926 B CN 108509926B
Authority
CN
China
Prior art keywords
color space
true
component
enhanced
extracting
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.)
Expired - Fee Related
Application number
CN201810307834.5A
Other languages
Chinese (zh)
Other versions
CN108509926A (en
Inventor
施文灶
程姗
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.)
Fujian Normal University
Original Assignee
Fujian Normal 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 Fujian Normal University filed Critical Fujian Normal University
Priority to CN201810307834.5A priority Critical patent/CN108509926B/en
Publication of CN108509926A publication Critical patent/CN108509926A/en
Application granted granted Critical
Publication of CN108509926B publication Critical patent/CN108509926B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/176—Urban or other man-made structures

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)

Abstract

The invention relates to a building extraction method based on bidirectional color space transformation. The method comprises the following steps: step 1, inputting a multispectral remote sensing image; step 2, converting the RGB color space into LUV color space; step 3, decomposing the LUV; step 4, extracting enhancement true sets in the components L, U and V respectively; step 5, converting the LUV color space into the RGB color space; step 6, image segmentation; step 7, post-treatment; and 8, outputting the result. The method can accurately extract the buildings in the multispectral remote sensing image, and can be applied to updating of the buildings in the urban geographical basic information database.

Description

Building extraction method based on bidirectional color space transformation
Technical Field
The invention relates to the field of remote sensing image processing, in particular to a building extraction method based on bidirectional color space transformation.
Background
The building is one of main geographic elements of a city and is important content of various city thematic maps, and the research on the extraction of the building has important significance for comprehensively investigating the city geographic information environment. With the rapid development of the high-resolution remote sensing image acquisition technology, the remote sensing image has better data sources for processing, analyzing and applying, and the digital product has wider and deeper application. The computer image processing technology, the pattern recognition, the artificial intelligence and the like all make progress to different degrees, and the possibility is provided for efficiently extracting effective information in massive images. However, the building information is much more difficult to extract than other information such as roads and water bodies, and the main reasons are as follows:
(1) the data source is mainly a two-dimensional remote sensing image, and direct three-dimensional data is lacked in most cases;
(2) different remote sensing images often have larger difference due to different factors such as spectral range, resolution, geometric images of the sensor, imaging conditions and the like;
(3) the appearances, texture details and the like of different types of buildings are varied, the differences on remote sensing images are large, a unified building model base is difficult to establish, and automatic extraction of information is difficult;
(4) the complexity of the scene of the building, such as low contrast, mutual shielding of houses, shadows of the building itself, shadows of other objects, and the like, makes it difficult to automatically extract the building with clear boundaries from the background.
Disclosure of Invention
The invention provides a building extraction method based on bidirectional color space transformation, which can overcome the problem of difficulty in extracting buildings in the current remote sensing image, can detect building targets in the remote sensing image by combining the advantages of RGB and LUV color spaces, does not need manual intervention, and has high automation degree.
The technical scheme adopted for realizing the aim of the invention is as follows: the method comprises the following steps:
step 1: inputting a multispectral remote sensing image I with width W and height H and containing three color components of R, G and Bin；
Step 2: the multispectral remote sensing image IinConverting from color space RGB to color space LUV;
and step 3: decomposing the color space LUV in the step 2 into a component L, a component U and a component V;
and 4, step 4: the components L of the color space LUV are converted to a true set (T) using the following formulaL) Indefinite set (X)L) And false set (P)L)：
Figure BDA0001621504380000021
In the formula (1), TL(x, y) is the true set (T)L) Value of the middle pixel (X, y), XL(X, y) is an indefinite set (X)L) Value of middle pixel point (x, y), PL(x, y) is a false set (P)L) The value of the middle pixel point (x, y); int (x, y) is the intensity value of the pixel (x, y),
Figure BDA0001621504380000022
is the average value of the intensity of all pixel points in a window with the side length of w multiplied by w and taking the pixel point (x, y) as the center,
Figure BDA0001621504380000023
is the maximum value of the average value of the intensities in the component L,
Figure BDA0001621504380000024
is the minimum of the average of the intensities in the component L; dif (x, y) is the intensity offset of the pixel point (x, y),
Figure BDA0001621504380000025
difgand are both intensity offset thresholds, and difg>difs；
And 5: the following formula is used to calculate the true set (T) in step 4L) Indefinite set (X)L) And false set (P)L) Respectively denoted as E (T)L)、E(XL) And E (P)L)：
Figure BDA0001621504380000026
Step 6: the information entropy E (T) in step 5 is combined using the following formulaL)、E(XL) And E (P)L) Obtaining the information entropy e (L) of the component L:
E(L)＝E(TL)+E(XL)+E(PL)； (3)
and 7: for the true set (T) in step 4L) Indefinite set (X)L) And false set (P)L) Performing enhancement operation, wherein the specific formula is as follows:
Figure BDA0001621504380000027
in the formula (4), TEN L、PEN LAnd XEN LRespectively, an enhanced true set, an enhanced false set and an enhanced false set in the component LGiven a set, k is an enhancement coefficient and can be calculated by the following formula:
Figure BDA0001621504380000031
ME is image IinThe maximum entropy of (D) is taken as log2 W×H；
And 8: judging the enhancement indeterminate set X in step 7 by the following formulaEN LInformation entropy E (X) ofEN L) Whether it is stable or not:
(E(XEN L)-E(XL))/E(XL)<δ， (6)
when equation (6) holds, the information entropy E (X)EN L) And (6) stabilizing, entering step 9, otherwise TL＝TEN L，XL＝XEN L，PL＝PEN LEntering step 5;
and step 9: extracting an enhanced true set TEN L；
Step 10: respectively extracting the enhanced true sets T in the components U by repeatedly using the methods from the step 4 to the step 9EN UEnhanced true set T in sum component VEN V；
Step 11: will enhance the true set TEN LEnhancing the true set TEN UAnd enhancing the true set TEN VThe formed LUV color space is converted into RGB color space to obtain image IT；
Step 12: for image ITCarrying out segmentation and extracting a segmentation object set SEG;
step 13: extracting the building by using the squareness and the length-width ratio as constraint conditions;
step 14: and outputting and extracting a building result.
The segmentation method in step 12 adopts a region growing segmentation method.
The invention has the beneficial effects that: the method can accurately extract the buildings in the multispectral remote sensing image, and can be applied to updating of the buildings in the urban geographical basic information database.
Drawings
FIG. 1 is an overall process flow diagram of the present invention;
fig. 2 is a flow chart of the present invention for extracting the enhanced proper set in the component L.
Detailed Description
The following detailed description of embodiments of the invention refers to the accompanying drawings.
FIG. 1 is a general process flow diagram of the present invention, as shown in FIG. 1:
in step 101, a multispectral remote sensing image I with width W and height H and containing three color components of R, G and B is inputin。
In step 102, the multispectral remote sensing image I in step 101 is processedinFrom the color space RGB to the color space LUV.
In step 103, the color space LUV in step 102 is decomposed into a component L, a component U, and a component V.
At step 104, the enhanced proper sets T in the components L, U and V are extracted separatelyEN LEnhancing the true set TEN UAnd enhancing the true set TEN V。
At step 105, the true set T will be enhancedEN LEnhancing the true set TEN UAnd enhancing the true set TEN VThe formed LUV color space is converted into RGB color space to obtain image IT。
In step 106, the region growing segmentation method is adopted to segment the image ITAnd (5) carrying out segmentation to extract a segmentation object set SEG.
At step 107, the building is extracted using the squareness and aspect ratio as constraints.
At step 108, the extracted building results are output.
FIG. 2 is a flow chart of the present invention for extracting the enhanced true set in the component L, i.e. the enhanced true set T included in step 104 in FIG. 1EN LThe extraction process is shown in figure 2:
in step 201, a component L is input.
In step 202, the components L of the color space LUV are converted to a proper set (T) using the following formulaL) Indefinite set (X)L) And false set (P)L)：
Figure BDA0001621504380000041
In the formula (7), TL(x, y) is the true set (T)L) Value of the middle pixel (X, y), XL(X, y) is an indefinite set (X)L) Value of middle pixel point (x, y), PL(x, y) is a false set (P)L) The value of the middle pixel point (x, y), int (x, y) is the intensity value of the pixel point (x, y),
Figure BDA0001621504380000042
is the average value of the intensity of all pixel points in a window with the side length of w multiplied by w and taking the pixel point (x, y) as the center,
Figure BDA0001621504380000043
is the maximum value of the average value of the intensities in the component L,
Figure BDA0001621504380000044
is the minimum value of the average value of the intensity in the component L, dif (x, y) is the intensity offset of the pixel point (x, y),
Figure BDA0001621504380000045
difgand are both intensity offset thresholds, and difg>difs。
In step 203, the true set (T) in step 202 is calculated by the following formulaL) Indefinite set (X)L) And false set (P)L) Respectively denoted as E (T)L)、E(XL) And E (P)L)：
Figure BDA0001621504380000046
In step 204, the step of merging is performed using the following formulaInformation entropy E (T) in 203L)、E(XL) And E (P)L) Obtaining the information entropy e (L) of the component L:
E(L)＝E(TL)+E(XL)+E(PL)。 (9)
in step 205, the true set (T) in step 202 is appliedL) Indefinite set (X)L) And false set (P)L) Performing enhancement operation, wherein the specific formula is as follows:
Figure BDA0001621504380000051
in the formula (10), TEN L、PEN LAnd XEN LRespectively, an enhanced true set, an enhanced false set and an enhanced indeterminate set in the component L, where k is an enhancement coefficient and can be calculated by the following formula:
Figure BDA0001621504380000052
ME is image IinThe maximum entropy of (D) is taken as log2 W×H。
In step 206, the enhancement indeterminate set X in step 205 is determined using the following formulaEN LInformation entropy E (X) ofEN L) Whether it is stable or not:
(E(XEN L)-E(XL))/E(XL)<δ， (12)
when equation (12) holds, the information entropy E (X)EN L) Stabilize, go to step 207, otherwise TL＝TEN L，XL＝XEN L，PL＝PEN LStep 203 is entered.
In step 207, an enhanced true set (T) of components L is outputEN L)。
With the process flow of FIG. 2, when component U is input at step 201, an enhanced truth set (T) for component U may be extractedEN U) When component V is input at step 201, it may beExtracting an enhanced true set (T) of components UEN V)。

Claims (2)

1. A building extraction method based on bidirectional color space transformation is characterized by comprising the following steps:
step 1: inputting a multispectral remote sensing image I with width W and height H and containing three color components of R, G and Bin；
Step 2: the multispectral remote sensing image IinConverting from color space RGB to color space LUV;
and step 3: decomposing the color space LUV in the step 2 into a component L, a component U and a component V;
and 4, step 4: the components L of the color space LUV are converted to a true set (T) using the following formulaL) Indefinite set (X)L) And false set (P)L)：
Figure FDA0003007920550000011
In the formula (1), TL(x, y) is the true set (T)L) Value of the middle pixel (X, y), XL(X, y) is an indefinite set (X)L) Value of middle pixel point (x, y), PL(x, y) is a false set (P)L) The value of the middle pixel point (x, y), int (x, y) is the intensity value of the pixel point (x, y),
Figure FDA0003007920550000012
is the average value of the intensity of all pixel points in a window with the side length of w multiplied by w and taking the pixel point (x, y) as the center,
Figure FDA0003007920550000013
is the maximum value of the average value of the intensities in the component L,
Figure FDA0003007920550000014
is the minimum value of the average value of the intensity in the component L, dif (x, y) is the intensity offset of the pixel point (x, y),
Figure FDA0003007920550000015
difgand difsAre all intensity offset thresholds, and difg>difs；
And 5: the following formula is used to calculate the true set (T) in step 4L) Indefinite set (X)L) And false set (P)L) Respectively denoted as E (T)L)、E(XL) And E (P)L)：
Figure FDA0003007920550000016
Step 6: the information entropy E (T) in step 5 is combined using the following formulaL)、E(XL) And E (P)L) Obtaining the information entropy e (L) of the component L:
E(L)＝E(TL)+E(XL)+E(PL)； (3)
and 7: for the true set (T) in step 4L) Indefinite set (X)L) And false set (P)L) Performing enhancement operation, wherein the specific formula is as follows:
Figure FDA0003007920550000021
in the formula (4), TEN L、PEN LAnd XEN LRespectively, an enhanced true set, an enhanced false set and an enhanced indeterminate set in the component L, where k is an enhancement coefficient and can be calculated by the following formula:
Figure FDA0003007920550000022
ME is image IinThe maximum entropy of (D) is taken as log2 W×H；
And 8: judging the enhancement indeterminate set X in step 7 by the following formulaEN LInformation entropy E of(XEN L) Whether it is stable or not:
(E(XEN L)-E(XL))/E(XL)<δ， (6)
when equation (6) holds, the information entropy E (X)EN L) And (6) stabilizing, entering step 9, otherwise TL＝TEN L，XL＝XEN L，PL＝PEN LEntering step 5;
and step 9: extracting an enhanced true set TEN L；
Step 10: respectively extracting the enhanced true sets T in the components U by repeatedly using the methods from the step 4 to the step 9EN UEnhanced true set T in sum component VEN V；
Step 11: will enhance the true set TEN LEnhancing the true set TEN UAnd enhancing the true set TEN VThe formed LUV color space is converted into RGB color space to obtain image IT；
Step 12: for image ITCarrying out segmentation and extracting a segmentation object set;
step 13: extracting the building by using the squareness and the length-width ratio as constraint conditions;
step 14: and outputting and extracting a building result.
2. The method for extracting buildings based on bidirectional color space transformation as claimed in claim 1, wherein the segmentation method in step 12 is a region growing segmentation method.
CN201810307834.5A 2018-04-08 2018-04-08 Building extraction method based on bidirectional color space transformation Expired - Fee Related CN108509926B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810307834.5A CN108509926B (en) 2018-04-08 2018-04-08 Building extraction method based on bidirectional color space transformation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810307834.5A CN108509926B (en) 2018-04-08 2018-04-08 Building extraction method based on bidirectional color space transformation

Publications (2)

Publication Number Publication Date
CN108509926A CN108509926A (en) 2018-09-07
CN108509926B true CN108509926B (en) 2021-06-01

Family

ID=63381209

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810307834.5A Expired - Fee Related CN108509926B (en) 2018-04-08 2018-04-08 Building extraction method based on bidirectional color space transformation

Country Status (1)

Country Link
CN (1) CN108509926B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105761250A (en) * 2016-02-01 2016-07-13 福建师范大学 Building extraction method based on fuzzy scene segmentation
CN107169946A (en) * 2017-04-26 2017-09-15 西北工业大学 Image interfusion method based on non-negative sparse matrix Yu hypersphere color transformation
CN107527007A (en) * 2016-06-20 2017-12-29 戴尔菲技术公司 For detecting the image processing system of perpetual object

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105761250A (en) * 2016-02-01 2016-07-13 福建师范大学 Building extraction method based on fuzzy scene segmentation
CN107527007A (en) * 2016-06-20 2017-12-29 戴尔菲技术公司 For detecting the image processing system of perpetual object
CN107169946A (en) * 2017-04-26 2017-09-15 西北工业大学 Image interfusion method based on non-negative sparse matrix Yu hypersphere color transformation

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
AUTOMATIC BUILDING DETECTION BASED ON CIE LUV COLOR SPACE USING VERY HIGH RESOLUTION PLEIADES IMAGES;Alireza Rahimzadeganasl 等;《IEEE》;20170629;全文 *
Building extraction from panchromatic high-resolution remotely sensed imagery based on potential histogram and neighborhood Total variation;Wenzao Shi 等;《Earth Sci Inform》;20160419;全文 *
基于颜色空间转换的交通图像增强算法;顾明 等;《仪器仪表学报》;20150831;第36卷(第8期);全文 *
高分辨率遥感图像建筑物提取方法研究;陈凯强;《中国优秀硕士学位论文全文数据库信息科技辑》;20170515(第5期);全文 *

Also Published As

Publication number Publication date
CN108509926A (en) 2018-09-07

Similar Documents

Publication Publication Date Title
Djurcilov et al. Visualizing scalar volumetric data with uncertainty
CN108876741B (en) Image enhancement method under complex illumination condition
Shen et al. Intrinsic image decomposition using optimization and user scribbles
CN106709964B (en) Sketch generation method and device based on gradient correction and multidirectional texture extraction
CN111611643A (en) Family type vectorization data obtaining method and device, electronic equipment and storage medium
CN104751432A (en) Image reconstruction based visible light and infrared image fusion method
CN111626951B (en) Image shadow elimination method based on content perception information
Zhao et al. Fast image enhancement using multi-scale saliency extraction in infrared imagery
CN114972107A (en) Low-illumination image enhancement method based on multi-scale stacked attention network
CN109635715B (en) Remote sensing image building extraction method
CN105574533A (en) Image feature extraction method and device
Huo et al. High‐dynamic range image generation from single low‐dynamic range image
CN103337073B (en) A kind of two dimensional image threshold segmentation method based on three-dimensional entropy
CN108509926B (en) Building extraction method based on bidirectional color space transformation
Zhang et al. Pulse coupled neural network edge-based algorithm for image text locating
CN108491826B (en) Automatic extraction method of remote sensing image building
CN106021610B (en) A kind of method for extracting video fingerprints based on marking area
Xiaoxu et al. Image dehazing base on two-peak channel prior
CN114299099A (en) Digital watermark extraction method, device, equipment and storage medium
Han et al. Automatic salient object segmentation using saliency map and color segmentation
CN108596088B (en) Building detection method for panchromatic remote sensing image
Sitdikov et al. Accelerated mutual entropy maximization for biomedical image registration
Dayal et al. Diagnostic utilisation of ground based imaging and non-imaging sensors for digital documentation of heritage sites
Zhou et al. Image enhancement and image matching of UAV based on histogram equalization
CN116778095B (en) Three-dimensional reconstruction method based on artificial intelligence

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20210601

CF01 Termination of patent right due to non-payment of annual fee