CN111383230A - Remote sensing image river extraction device and method - Google Patents
Remote sensing image river extraction device and method Download PDFInfo
- Publication number
- CN111383230A CN111383230A CN201811613683.2A CN201811613683A CN111383230A CN 111383230 A CN111383230 A CN 111383230A CN 201811613683 A CN201811613683 A CN 201811613683A CN 111383230 A CN111383230 A CN 111383230A
- Authority
- CN
- China
- Prior art keywords
- water body
- body index
- river
- parameters
- river extraction
- 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
- 238000000605 extraction Methods 0.000 title claims abstract description 60
- 238000000034 method Methods 0.000 title claims abstract description 35
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 62
- 230000000903 blocking effect Effects 0.000 claims abstract description 30
- 238000004364 calculation method Methods 0.000 claims abstract description 17
- 230000011218 segmentation Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 abstract description 21
- 238000005192 partition Methods 0.000 description 3
- 230000004927 fusion Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000002310 reflectometry Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000015654 memory Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
Images
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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
Landscapes
- Engineering & Computer Science (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 remote sensing image river extraction device and method. The device includes: the main process unit is used for receiving the remote sensing image, blocking the image according to the size of the specified blocks, generating blocking parameters and distributing the blocking parameters to the plurality of computing units; and the plurality of calculation units are used for receiving the blocking parameters, calculating a river extraction result according to the blocking parameters and outputting the river extraction result. The NDWI river extraction algorithm is modified by the device and the method based on the MPI (message serving interface) distributed framework, so that the normalized difference water body index distributed processing is realized, the dependence on resources is greatly reduced, and the processing speed of river extraction in the remote sensing image is improved.
Description
Technical Field
The invention relates to the technical field of remote sensing, in particular to a device and a method for extracting a river from a remote sensing image.
Background
The Normalized Difference Water Index (NDWI) is a common river extraction method, and the idea is that in the near infrared, mid infrared and short wave infrared parts, the Water body almost absorbs all incident energy, so that the reflectivity of the Water body in these bands is particularly low, and the energy absorbed by soil, vegetation, buildings and the like in these bands is small and has high reflectivity, so that the Water body is obviously different from the bands. The water body index is constructed by utilizing the characteristics, and can be used for distinguishing the water body from other ground objects. The method is mainly realized by applying a plug-in mode to single-machine software.
With the spatial resolution of the remote sensing image being improved to a meter level or even a sub-meter level, the data volume of the multispectral image after single scene fusion reaches dozens of G, and if the multispectral image after single scene fusion is embedded among multiple scenes, the data volume can reach hundreds of G. However, when rivers are distributed around the world, in order to extract the rivers in the region of interest, all the images of the rivers must be processed once, which requires a large amount of data to be processed. Such data volume cannot meet the requirement of rapid processing completion depending on single machine processing; however, in the current distributed processing implementation mode, a control center is generally arranged on the upper layer, a plurality of processing units are arranged on the lower layer, processing tasks are decomposed on the plurality of processing units, and in order to improve processing performance, the memories, processors and the like of the processing units need to be upgraded, so that it is seen that the existing normalized difference water body index method is seriously affected by operating environment resources, and remote sensing images, particularly remote sensing image water body calculation with large data volume and river extraction speed improvement are hindered. In addition, due to the limitation that only one file can be operated in one process, the result of the water body calculation can be output to one file only in sequence, and the processing time is increased indirectly.
Disclosure of Invention
The invention aims to solve at least one of the technical problems in the prior art or the related art and provides a scheme capable of improving the river extraction speed.
To this end, according to a first aspect of the present invention, there is provided a remote sensing image river extraction device, comprising:
the main process unit is used for receiving the remote sensing image, blocking the image according to the size of the specified blocks, generating blocking parameters and distributing the blocking parameters to the plurality of computing units;
and the plurality of calculation units are used for receiving the blocking parameters, calculating a river extraction result according to the blocking parameters and outputting the river extraction result.
Further, the calculating the river extraction result according to the blocking parameters comprises:
calculating according to the blocking parameters by using a normalized difference water body index formula to obtain a water body index;
and (4) performing threshold segmentation on the water body index, and determining that the water body index is a river if the water body index is greater than a specific threshold.
Further, the normalized difference water body index formula is as follows:
wherein, green is the pixel value of the green wave band, nir is the pixel value of the near infrared wave band, and NDWI is the water body index obtained by calculation.
Further, characterized in that the specific threshold is 0.3.
Further, the device further comprises a read-write library supporting parallel, and the outputting the river extraction result comprises:
outputting the river extraction result to the read-write library in parallel in a form of 'header file + block binary file', wherein the header file records the block output of the divided images, the merging sequence of the output blocks and the overall description information of the output result; and recording the river extraction result of the corresponding computing unit by the block binary file.
According to a second aspect of the invention, a remote sensing image river extraction method is provided, and is characterized by comprising the following steps:
the method comprises the steps that a main process unit receives a remote sensing image, blocks the image according to the size of a specified block, generates block parameters and distributes the block parameters to a plurality of computing units;
the calculation unit receives the block parameters, calculates a river extraction result according to the block parameters and outputs the river extraction result.
Further, the calculating the river extraction result according to the blocking parameters comprises:
calculating according to the blocking parameters by using a normalized difference water body index formula to obtain a water body index;
and (4) performing threshold segmentation on the water body index, and determining that the water body index is a river if the water body index is greater than a specific threshold.
Further, the normalized difference water body index formula is as follows:
wherein, green is the pixel value of the green wave band, nir is the pixel value of the near infrared wave band, and NDWI is the water body index obtained by calculation.
Further, the specific threshold is 0.3.
Further, the outputting the river extraction result comprises:
outputting river extraction results in parallel to a read-write library supporting parallel through a head file + block binary file mode, wherein the head file records the block output of the divided images, the merging sequence of the output blocks and the integral description information of the output results; and recording the river extraction result of the corresponding computing unit by the block binary file.
The invention provides a device and a method based on an MPI (message serving interface) distributed framework, which are used for modifying an NDWI river extraction algorithm, realizing the normalized difference water body index distributed processing, greatly reducing the dependence on resources and improving the processing speed of river extraction in remote sensing images. In addition, the improved data parallel read-write library is utilized to provide parallelization of the whole processing flow, and the output speed of river extraction results is further improved.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a structural view of a river extraction device according to the present invention;
fig. 2 is a flowchart of a river extraction method according to the present invention.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described herein, and therefore the scope of the present invention is not limited by the specific embodiments disclosed below.
The MPI framework provides a low-latency computing task execution service for a set of similar computing requirements. The framework contains two roles: a main process and a computing unit. The main process is responsible for receiving requests from external programs, and the computing unit is responsible for initiating and monitoring specific computing tasks. The computing units are isomorphic computing units, and each computing unit can only accommodate one computing task at the same time. According to the invention, through the MPI framework-based device, the distributed processing and parallelization output of the NDWI river extraction algorithm are realized, and the algorithm performance is improved.
Referring to fig. 1, there is shown the structure of a river diversion apparatus according to the present invention. The device comprises a main process unit P1 and a plurality of computing units C1, C2, C3, M (M is a natural number greater than 1), wherein the main process unit P1 is used for receiving a remote sensing image, blocking the image according to a specified blocking size, generating blocking parameters and distributing the blocking parameters to the plurality of computing units, and the remote sensing image is optionally a multispectral remote sensing image; the plurality of calculation units C1, C2, C3, a.
Further, the receiving and processing the water body computing subtask comprises:
calculating according to the blocking parameters by using a normalized difference water body index formula to obtain a water body index;
optionally, the formula is:
wherein, green is a pixel value of a green wave band, nir is a pixel value of a near-infrared wave band, and NDWI is a water body index obtained by calculation;
and after the water body index is obtained, performing threshold segmentation on the water body index, and if the water body index is larger than a specific threshold, determining that the water body index is a river. Optionally, the specific threshold is 0.3.
If the results of the computing units are output in sequence while the processing process is parallelized, the time of the output process is still long, and the parallelization processing speed of the algorithm is slowed down due to the fact that the computing units wait for output in a queue. Therefore, the device further comprises a read-write library S1 supporting parallel under the MPI framework, and each computing unit C1, C2, C3, ·. · and CM outputs the river extraction result in parallel to the read-write library in the form of "header file + block binary file". Specifically, when outputting the river extraction result, a header file is first written out, which mainly records how many blocks (one block corresponds to one calculation unit) the input image is divided into in total to be output, the merging order of the output blocks, and the overall description information of the output result. Secondly, according to a parallel mode, each computing unit writes out respective binary files at the same time, and river extraction results of the corresponding computing units are recorded in the binary files. Thus, the complete river extraction result is stored in the header file and the block binary file.
Referring to fig. 2, there is shown a flow of the river extraction method according to the present invention. The invention reforms NDWI river extraction algorithm according to the design idea of MPI distributed framework, the method includes the steps:
s21, receiving the remote sensing image by the main process unit, wherein the remote sensing image is optionally a multispectral remote sensing image;
s22, partitioning the image according to the specified partition size, generating partition parameters and distributing the partition parameters to a plurality of computing units;
s23, the calculation unit receives the block parameters and calculates river extraction results according to the block parameters;
in step S23, calculating a river extraction result from the blocking parameter includes:
calculating according to the blocking parameters by using a normalized difference water body index formula to obtain a water body index;
optionally, the normalized difference water body index formula is:
wherein, green is the pixel value of the green wave band, nir is the pixel value of the near infrared wave band, and NDWI is the water body index obtained by calculation.
And (4) performing threshold segmentation on the water body index, and determining that the water body index is a river if the water body index is greater than a specific threshold. Optionally, the specific threshold is 0.3.
S24, outputting the river extraction result;
the method comprises the following steps: and outputting the river extraction result in parallel to a read-write library supporting parallel through a head file + block binary file form.
The device and the method based on the MPI (message serving interface) distributed framework realize the distributed processing of the normalized difference water body index, greatly reduce the dependence on resources and improve the processing speed of river extraction in the remote sensing image; meanwhile, the parallel read-write library is utilized to provide parallelization of the whole processing flow, and the output speed of the river extraction result is further improved.
A detailed process of river extraction is set forth below in one embodiment of the invention, including:
a. inputting a multispectral remote sensing image;
b. the main process unit receives the remote sensing image;
c. the main process unit divides the image into blocks according to the size of the specified blocks, generates block parameters and distributes the block parameters to a plurality of computing units, and the block parameters to be processed by each computing unit are different, so that the computing can be performed in parallel; the blocking parameters comprise a green-band pixel value green and a near-infrared-band pixel value nir;
d. each calculating unit receives corresponding block parameters;
e. each calculating unit calculates and obtains a water body index according to the corresponding block parameters by using a normalized difference water body index formula;
optionally, the normalized difference water body index formula is:
the blocking parameters comprise green and nir, the green is a pixel value of a green band, the nir is a pixel value of a near-infrared band, and the NDWI is a water body index obtained by calculation.
f. Judging whether the water body index is greater than 0.3, if so, determining that the water body index is a river image, otherwise, determining that the water body index is a non-river image;
g. each computing unit generates a river extraction result according to the judgment of whether the river image is the river image;
h. and outputting river extraction results to a read-write library in parallel. The method comprises the following steps:
firstly, writing a header file, wherein the header file mainly records the output of how many blocks of input images are divided into, the merging sequence of output blocks and the overall description information of output results;
secondly, according to a parallel mode, each computing unit writes out respective binary files at the same time, and river extraction results of the corresponding computing units are recorded in the binary files.
Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium. The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A remote sensing image river extraction device is characterized by comprising:
the main process unit is used for receiving the remote sensing image, blocking the image according to the size of the specified blocks, generating blocking parameters and distributing the blocking parameters to the plurality of computing units;
and the plurality of calculation units are used for receiving the blocking parameters, calculating a river extraction result according to the blocking parameters and outputting the river extraction result.
2. The apparatus of claim 1, wherein said calculating river extraction results from said chunking parameters comprises:
calculating according to the blocking parameters by using a normalized difference water body index formula to obtain a water body index;
and (4) performing threshold segmentation on the water body index, and determining that the water body index is a river if the water body index is greater than a specific threshold.
3. The apparatus of claim 2, wherein the normalized difference water body index formula is:
wherein, green is the pixel value of the green wave band, nir is the pixel value of the near infrared wave band, and NDWI is the water body index obtained by calculation.
4. The apparatus of claim 3, wherein the specific threshold is 0.3.
5. The apparatus according to any one of claims 1-4, wherein the apparatus further comprises a read-write library supporting parallel, and the outputting the river extraction result comprises:
outputting the river extraction result to the read-write library in parallel in a form of 'header file + block binary file', wherein the header file records the block output of the divided images, the merging sequence of the output blocks and the overall description information of the output result; and recording the river extraction result of the corresponding computing unit by the block binary file.
6. A remote sensing image river extraction method is characterized by comprising the following steps:
the method comprises the steps that a main process unit receives a remote sensing image, blocks the image according to the size of a specified block, generates block parameters and distributes the block parameters to a plurality of computing units;
the calculation unit receives the block parameters, calculates a river extraction result according to the block parameters and outputs the river extraction result.
7. The method of claim 6, wherein said calculating river extraction results from said chunking parameters comprises:
calculating according to the blocking parameters by using a normalized difference water body index formula to obtain a water body index;
and (4) performing threshold segmentation on the water body index, and determining that the water body index is a river if the water body index is greater than a specific threshold.
8. The method of claim 7, wherein the normalized difference water body index formula is:
wherein, green is the pixel value of the green wave band, nir is the pixel value of the near infrared wave band, and NDWI is the water body index obtained by calculation.
9. The method of claim 8, wherein the specific threshold is 0.3.
10. The method according to any one of claims 6-9, wherein the outputting the river extraction result comprises:
outputting river extraction results in parallel to a read-write library supporting parallel through a head file + block binary file mode, wherein the head file records the block output of the divided images, the merging sequence of the output blocks and the integral description information of the output results; and recording the river extraction result of the corresponding computing unit by the block binary file.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811613683.2A CN111383230A (en) | 2018-12-27 | 2018-12-27 | Remote sensing image river extraction device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811613683.2A CN111383230A (en) | 2018-12-27 | 2018-12-27 | Remote sensing image river extraction device and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111383230A true CN111383230A (en) | 2020-07-07 |
Family
ID=71214669
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811613683.2A Pending CN111383230A (en) | 2018-12-27 | 2018-12-27 | Remote sensing image river extraction device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111383230A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112102288A (en) * | 2020-09-15 | 2020-12-18 | 北京百度网讯科技有限公司 | Water body identification and water body change detection method, device, equipment and medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104915954A (en) * | 2015-05-25 | 2015-09-16 | 同济大学 | Automatic water body extraction method based on Landsat OLI multispectral remote sensing image |
| CN105046087A (en) * | 2015-08-04 | 2015-11-11 | 中国资源卫星应用中心 | Water body information automatic extraction method for multi-spectral image of remote sensing satellite |
| CN108985209A (en) * | 2018-07-06 | 2018-12-11 | 航天星图科技（北京）有限公司 | A kind of remote sensing image urban green space extracting method based on Distributed Architecture |
| CN109117722A (en) * | 2018-07-06 | 2019-01-01 | 航天星图科技（北京）有限公司 | A kind of remote sensing image river extracting method based on Distributed Architecture |
-
2018
- 2018-12-27 CN CN201811613683.2A patent/CN111383230A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104915954A (en) * | 2015-05-25 | 2015-09-16 | 同济大学 | Automatic water body extraction method based on Landsat OLI multispectral remote sensing image |
| CN105046087A (en) * | 2015-08-04 | 2015-11-11 | 中国资源卫星应用中心 | Water body information automatic extraction method for multi-spectral image of remote sensing satellite |
| CN108985209A (en) * | 2018-07-06 | 2018-12-11 | 航天星图科技（北京）有限公司 | A kind of remote sensing image urban green space extracting method based on Distributed Architecture |
| CN109117722A (en) * | 2018-07-06 | 2019-01-01 | 航天星图科技（北京）有限公司 | A kind of remote sensing image river extracting method based on Distributed Architecture |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112102288A (en) * | 2020-09-15 | 2020-12-18 | 北京百度网讯科技有限公司 | Water body identification and water body change detection method, device, equipment and medium |
| CN112102288B (en) * | 2020-09-15 | 2023-11-07 | 应急管理部大数据中心 | Water body identification and water body change detection method, device, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Matsubara et al. | Neural compression and filtering for edge-assisted real-time object detection in challenged networks | |
| KR102192211B1 (en) | Efficient Generative Adversarial Networks using Depthwise Separable and Channel Attention for Image to Image Translation | |
| WO2017005149A1 (en) | Game acceleration method and device | |
| WO2021129466A1 (en) | Watermark detection method, device, terminal and storage medium | |
| Song et al. | Extraction of loess shoulder-line based on the parallel GVF snake model in the loess hilly area of China | |
| CN111753721A (en) | Human body posture recognition method and device | |
| CN104574358A (en) | Method and apparatus for scene segmentation from focal stack images | |
| CN105931256A (en) | CUDA (compute unified device architecture)-based large-format remote sensing image fast segmentation method | |
| US20220398712A1 (en) | Generating modified digital images utilizing nearest neighbor fields from patch matching operations of alternate digital images | |
| CN110176021A (en) | In conjunction with the level set image segmentation method and system of the conspicuousness information of gamma correction | |
| KR102082293B1 (en) | Device and method for binarization computation of convolution neural network | |
| CN111383230A (en) | Remote sensing image river extraction device and method | |
| CN107920253B (en) | Video processing method based on GPU | |
| US9232222B2 (en) | Lossless color image compression adaptively using spatial prediction or inter-component prediction | |
| CN111340790A (en) | Bounding box determination method and device, computer equipment and storage medium | |
| US20140133772A1 (en) | Electronic device and method for splitting image | |
| CN104657733B (en) | A kind of massive medical image data memory storage and method | |
| RU2671304C1 (en) | Method and system for constructing digital print of video content | |
| CN114201646A (en) | Video retrieval method and system | |
| KR101212845B1 (en) | Method And System For Sampling Moving Picture | |
| KR20100009451A (en) | Method for determining ground line | |
| CN106331720B (en) | Video decoding related information storage method and device | |
| AU2009212933A1 (en) | Methods of storing and retrieving images | |
| CN105550971B (en) | A kind of vector geographic data security level identification method based on distance ratio modulation | |
| CN110502994B (en) | Normalized vegetation index parallel extraction method based on OpenACC |
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 |