CN112990001A - Remote sensing image monomer building rapid and accurate extraction method applied to digital collection - Google Patents
Remote sensing image monomer building rapid and accurate extraction method applied to digital collection Download PDFInfo
- Publication number
- CN112990001A CN112990001A CN202110266481.0A CN202110266481A CN112990001A CN 112990001 A CN112990001 A CN 112990001A CN 202110266481 A CN202110266481 A CN 202110266481A CN 112990001 A CN112990001 A CN 112990001A
- Authority
- CN
- China
- Prior art keywords
- building
- remote sensing
- sensing image
- model
- image
- 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
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
Abstract
The invention relates to a method for quickly and accurately extracting a remote sensing image monomer building applied to digital sign, which is characterized by comprising the following steps of: s1, obtaining a remote sensing image in the engineering range; s2, obtaining a training sample set; s3, obtaining a trained building model; s4, taking the remote sensing image in the engineering range as input, and predicting by adopting a trained building model to obtain gridded building data in the remote sensing image, wherein the value range of each grid in the building data is a first threshold range, the maximum value represents that the highest probability belongs to a building, and the minimum value represents that the highest probability belongs to a non-building; s5, when the missing extraction and the false extraction exist in the building data, obtaining a building sample file added aiming at the missing extraction and the false extraction area, updating the training sample set, and repeating the steps S2 to S4 until the predicted building data meet the requirements; and S6, converting the rasterized building data into a vector image spot result of the building in a raster-to-vector mode.
Description
Technical Field
The invention relates to a method for quickly and accurately extracting a remote sensing image monomer building applied to digital collection. The method is suitable for the field of high-resolution remote sensing image information extraction.
Background
The expropriation is a process of expropriating the land by the government according to the requirements of national economy and social development planning, general land utilization planning, urban and rural planning, special planning and the like, wherein the problem of house removal expropriation of the land is involved, so that the current situation of buildings in the expropriation range needs to be acquired, and then the information of expropriated residents is associated with the building pattern spots to provide guidance for the subsequent specific expropriation process.
The building is an important monitoring target element of land utilization and earth surface coverage change, roof information of the building can be clearly distinguished in remote sensing images, and the situations such as the number, the area, the distribution and the like of the building can be accurately reflected.
In recent years, the method for extracting buildings from high-resolution remote sensing images is widely applied due to the fact that the building image spots can be acquired more quickly, and secondly, the method for extracting buildings from remote sensing images is continuously updated, and the method is developed into a semantic segmentation technology based on a deep convolutional network which is more effective at present from a pixel-based multi-scale segmentation technology and an object-oriented classification technology, and the current popular networks include FCN, U-Net, PSPNet, Mask-RCNN, D-LinkNet and the like.
However, in the application of digital migration, not only the building extraction, but also the selected cutting of the pre-extraction migration range of the building, the post-processing of the building image spots after the building extraction, and the calibration of attributes such as the user information of the building migration are often involved.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: aiming at the existing problems, the method for rapidly and accurately extracting the remote sensing image monomer building applied to digital sign is provided.
The technical scheme adopted by the invention is as follows: a remote sensing image monomer building rapid and accurate extraction method applied to digital collection is characterized by comprising the following steps:
s1, obtaining a remote sensing image in the engineering range;
s2, obtaining a training sample set, wherein the training sample set comprises a plurality of rasterized building sample files, the building sample files are formed by grid-converting operation of image samples cut from the remote sensing images and containing building images, and building outlines are drawn on the image samples and marked with buildings and background labels;
s3, training by adopting a convolutional neural network model according to the training sample set to obtain a trained building model;
s4, taking the remote sensing image in the engineering range as input, and predicting by adopting a trained building model to obtain gridded building data in the remote sensing image, wherein the value range of each grid in the building data is a first threshold range, the maximum value represents that the highest probability belongs to a building, and the minimum value represents that the highest probability belongs to a non-building;
s5, when the missing extraction and the false extraction exist in the building data, obtaining a building sample file added aiming at the missing extraction and the false extraction area, updating the training sample set, and repeating the steps S2 to S4 until the predicted building data meet the requirements;
and S6, converting the rasterized building data into a vector image spot result of the building in a raster-to-vector mode.
The step S1 includes:
and obtaining a remote sensing image of the latest time phase and an engineering range surface vector file drawn according to the remote sensing image, and cutting the remote sensing image by adopting the surface vector file to obtain the remote sensing image in the engineering range.
The image sample has a length of 600 to 1000 pixels and a width of 600 to 1000 pixels.
After the training sample set is updated in step S5, model training is continued with the previously trained building model as a pre-model.
The utility model provides a be applied to quick accurate extraction element of remote sensing image monomer building of digit expropriation, its characterized in that:
the image acquisition module is used for acquiring a remote sensing image in an engineering range;
the system comprises a sample acquisition module, a data processing module and a data processing module, wherein the sample acquisition module is used for acquiring a training sample set, the training sample set comprises a plurality of rasterized building sample files, the building sample files are formed by converting image samples cut from the remote sensing images and containing building images into grids, and the building outlines are drawn on the image samples and marked with buildings and background labels;
the model training module is used for training by adopting a convolutional neural network model according to a training sample set to obtain a trained building model;
the model prediction module is used for taking the remote sensing image in the engineering range as input and adopting a trained building model to carry out prediction to obtain gridded building data in the remote sensing image, wherein the value range of each grid in the building data is a first threshold value range, the maximum value represents that the highest probability belongs to a building, and the minimum value represents that the highest probability belongs to a non-building;
the updating iteration module is used for acquiring building sample files added aiming at the missed-lifting and false-lifting areas when the missed-lifting and false-lifting exist in the building data, and updating the training sample set until the predicted building data meet the requirements;
and the vector map conversion module is used for converting the rasterized building data into a vector map spot result of the building in a raster-to-vector mode.
The remote sensing image in the engineering scope is obtained by the following steps:
and obtaining a remote sensing image of the latest time phase and an engineering range surface vector file drawn according to the remote sensing image, and cutting the remote sensing image by adopting the surface vector file to obtain the remote sensing image in the engineering range.
The image sample has a length of 600 to 1000 pixels and a width of 600 to 1000 pixels.
And after the training sample set is updated in the updating iteration module, model training is continuously carried out by taking the previously trained building model as a pre-model.
A storage medium having stored thereon a computer program executable by a processor, the computer program comprising: and when being executed, the computer program realizes the steps of the method for quickly and accurately extracting the remote sensing image monomer building applied to digital migration.
A computer device having a memory and a processor, the memory having stored thereon a computer program executable by the processor, the computer program comprising: and when being executed, the computer program realizes the steps of the method for quickly and accurately extracting the remote sensing image monomer building applied to digital migration.
The invention has the beneficial effects that: the method takes the remote sensing image in the engineering range as input, adopts the building model to predict, obtains the gridded building data in the remote sensing image, and converts the building data into the vector graphic spots of the building, thereby realizing the rapid and accurate extraction of the single building.
The building model is obtained by training a training sample set formed by converting image samples cut from a remote sensing image in an engineering range into grids, is more suitable for predicting the building on the remote sensing image, and optimizes the building model by pertinently increasing the training samples when the predicted building data cannot meet the requirements.
Drawings
FIG. 1 is a flow chart of the present invention.
FIG. 2 is a logic flow diagram of the present invention.
Fig. 3 is a flowchart of the device for rapidly and accurately extracting the remote sensing image single building.
Fig. 4 is a schematic diagram illustrating the effect of the present invention.
Fig. 5 is a remote sensing image map used as an example of the present invention.
Fig. 6 is a vector diagram spot calculated by the present invention.
Fig. 7 is a final vector diagram.
FIG. 8 is a vector diagram after vector beautification of the remote sensing image monomer building rapid accurate extraction device.
In the figure: 1. the device comprises an image acquisition module 2, a sample acquisition module 3, a model training module 4, a model prediction module 5, an update iteration module 6 and a vector diagram conversion module.
Detailed Description
The embodiment is a method for quickly and accurately extracting a remote sensing image monomer building applied to digital sign, and the method specifically comprises the following steps:
and S1, obtaining a high-resolution remote sensing image of the latest time phase, obtaining a face vector file of a digital expropriation engineering range obtained by manual drawing of a worker according to the remote sensing image, determining a target range vector extracted by the building, and cutting the remote sensing image by adopting the face vector file to obtain the remote sensing image in the engineering range.
S2, obtaining a training sample set, wherein the training sample set comprises a plurality of rasterized building sample files, the building sample files are formed by converting image samples cut from remote sensing images in an engineering range and containing building images into grids, and building outlines are drawn on the image samples and marked with buildings and background labels.
In this embodiment, a worker selects a representative building area based on an obtained remote sensing image in an engineering range, covers various building types as much as possible, and cuts the remote sensing image to obtain a plurality of samples including a corresponding raster image and a plane vector file, wherein the sample size is X pixels, the width is Y pixels, and generally, the preferred value range of X, Y is 600-1000 pixels; and then, further labeling a corresponding face vector file according to each sample raster image, manually drawing the outline of the building, labeling the building and a background label, converting the labeled vector data into raster data to obtain a rasterized building sample file, namely completing the drawing of the building sample, and forming a training sample set.
And S3, training by adopting a convolutional neural network model according to the training sample set to obtain a trained building model.
S4, taking the remote sensing image in the engineering range as input, and adopting the trained building model to predict to obtain gridded building data in the remote sensing image, namely a strength map of the building, wherein the strength map is a grid map with the same size as the target area image, each grid value range is a first threshold value range, [0-255], wherein 255 represents that the highest probability belongs to the building, and 0 represents that the highest probability belongs to the non-building.
And S5, when the missing lifting and false lifting exist in the building data, acquiring a building sample file which is added in a proper amount aiming at the missing lifting and false lifting area, updating the training sample set, training by using the new sample set and using the model trained last time as a pre-model, predicting images, and iterating for multiple times until a satisfactory result is obtained (for example, the missing lifting and false lifting does not exist in the building data).
And S6, converting the rasterized building data into a vector image spot result of the building in a raster-to-vector mode.
S7, the building boundary of the building vector image spot result obtained by model training presents irregular boundaries such as saw-tooth shape and the like, and is not smooth enough, so that post-processing is needed, and in the embodiment, the building vector result which is more beautiful and practical is obtained by smoothing, straightening, removing inflection points and the like on the vectorized building image spot.
S8, the building further relates to information of the resident to each building in the digital migration application, and it is necessary to associate the building pattern spot with the information of the resident to assign the information of the corresponding resident to each building, and the obtained worker adds corresponding attributes to each building pattern spot based on the information of the resident to each building in this embodiment.
The embodiment also provides a rapid and accurate extraction device for the remote sensing image single building applied to digital collection, which comprises an image acquisition module, a sample acquisition module, a model training module, a model prediction module, an updating iteration module, a vector diagram conversion module, a vector diagram beautifying module and an attribute adding module.
The image acquisition module is used for acquiring remote sensing images in an engineering range; the sample acquisition module is used for acquiring a training sample set, the training sample set comprises a plurality of rasterized building sample files, the building sample files are formed by converting image samples cut from the remote sensing images and containing building images into grids, and building outlines are drawn on the image samples and marked with buildings and background labels; the model training module is used for training by adopting a convolutional neural network model according to a training sample set to obtain a trained building model; the model prediction module is used for taking the remote sensing image in the engineering range as input and adopting a trained building model to carry out prediction to obtain gridded building data in the remote sensing image, the value range of each grid in the building data is a first threshold value range, wherein the maximum value represents that the highest probability belongs to a building, and the minimum value represents that the highest probability belongs to a non-building; the updating iteration module is used for acquiring a building sample file added aiming at an extraction missing and false extraction area when the extraction missing and false extraction exists in the building data, and updating the training sample set until the predicted building data meets the requirement; the vector diagram conversion module is used for converting the rasterized building data into a vector diagram spot result of the building in a grid-to-vector mode; the vector diagram beautifying module is used for obtaining a more beautiful and practical building vector result by performing operations such as smoothing, straightening, inflection point removal and the like on the vectorized building diagram; and the attribute adding module is used for acquiring information of the expropriated residents and adding corresponding attributes to each building pattern spot by the staff.
The present embodiment also provides a storage medium, on which a computer program executable by a processor is stored, where the computer program is executed to implement the steps of the method for fast and accurate extraction of a single remote sensing image building applied to digital migration in the present embodiment.
The embodiment also provides a computer device, which has a memory and a processor, wherein the memory stores a computer program executable by the processor, and the computer program implements the steps of the method for rapidly and accurately extracting the remote sensing image monomer building applied to digital representation in the embodiment when being executed.
Claims (10)
1. A remote sensing image monomer building rapid and accurate extraction method applied to digital collection is characterized by comprising the following steps:
s1, obtaining a remote sensing image in the engineering range;
s2, obtaining a training sample set, wherein the training sample set comprises a plurality of rasterized building sample files, the building sample files are formed by grid-converting operation of image samples cut from the remote sensing images and containing building images, and building outlines are drawn on the image samples and marked with buildings and background labels;
s3, training by adopting a convolutional neural network model according to the training sample set to obtain a trained building model;
s4, taking the remote sensing image in the engineering range as input, and predicting by adopting a trained building model to obtain gridded building data in the remote sensing image, wherein the value range of each grid in the building data is a first threshold range, the maximum value represents that the highest probability belongs to a building, and the minimum value represents that the highest probability belongs to a non-building;
s5, when the missing extraction and the false extraction exist in the building data, obtaining a building sample file added aiming at the missing extraction and the false extraction area, updating the training sample set, and repeating the steps S2 to S4 until the predicted building data meet the requirements;
and S6, converting the rasterized building data into a vector image spot result of the building in a raster-to-vector mode.
2. The method for rapidly and accurately extracting the remote sensing image monomer building applied to digital migration according to claim 1, wherein the step S1 comprises:
and obtaining a remote sensing image of the latest time phase and an engineering range surface vector file drawn according to the remote sensing image, and cutting the remote sensing image by adopting the surface vector file to obtain the remote sensing image in the engineering range.
3. The method for rapidly and accurately extracting the remote sensing image monomer building applied to digital migration according to claim 1, characterized in that: the image sample has a length of 600 to 1000 pixels and a width of 600 to 1000 pixels.
4. The method for rapidly and accurately extracting the remote sensing image monomer building applied to digital migration according to claim 1, characterized in that: after the training sample set is updated in step S5, model training is continued with the previously trained building model as a pre-model.
5. The utility model provides a be applied to quick accurate extraction element of remote sensing image monomer building of digit expropriation, its characterized in that:
the image acquisition module is used for acquiring a remote sensing image in an engineering range;
the system comprises a sample acquisition module, a data processing module and a data processing module, wherein the sample acquisition module is used for acquiring a training sample set, the training sample set comprises a plurality of rasterized building sample files, the building sample files are formed by converting image samples cut from the remote sensing images and containing building images into grids, and the building outlines are drawn on the image samples and marked with buildings and background labels;
the model training module is used for training by adopting a convolutional neural network model according to a training sample set to obtain a trained building model;
the model prediction module is used for taking the remote sensing image in the engineering range as input and adopting a trained building model to carry out prediction to obtain gridded building data in the remote sensing image, wherein the value range of each grid in the building data is a first threshold value range, the maximum value represents that the highest probability belongs to a building, and the minimum value represents that the highest probability belongs to a non-building;
the updating iteration module is used for acquiring building sample files added aiming at the missed-lifting and false-lifting areas when the missed-lifting and false-lifting exist in the building data, and updating the training sample set until the predicted building data meet the requirements;
and the vector map conversion module is used for converting the rasterized building data into a vector map spot result of the building in a raster-to-vector mode.
6. The device for rapidly and accurately extracting the remote sensing image monomer building applied to digital migration according to claim 5, wherein the obtaining of the remote sensing image in the engineering scope comprises:
and obtaining a remote sensing image of the latest time phase and an engineering range surface vector file drawn according to the remote sensing image, and cutting the remote sensing image by adopting the surface vector file to obtain the remote sensing image in the engineering range.
7. The device for rapidly and accurately extracting the remote sensing image monomer building applied to digital migration according to claim 5, is characterized in that: the image sample has a length of 600 to 1000 pixels and a width of 600 to 1000 pixels.
8. The device for rapidly and accurately extracting the remote sensing image monomer building applied to digital migration according to claim 5, is characterized in that: and after the training sample set is updated in the updating iteration module, model training is continuously carried out by taking the previously trained building model as a pre-model.
9. A storage medium having stored thereon a computer program executable by a processor, the computer program comprising: the computer program is used for realizing the steps of the method for rapidly and accurately extracting the remote sensing image monomer building applied to digital migration in any one of claims 1 to 4 when being executed.
10. A computer device having a memory and a processor, the memory having stored thereon a computer program executable by the processor, the computer program comprising: the computer program is used for realizing the steps of the method for rapidly and accurately extracting the remote sensing image monomer building applied to digital migration in any one of claims 1 to 4 when being executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110266481.0A CN112990001A (en) | 2021-03-11 | 2021-03-11 | Remote sensing image monomer building rapid and accurate extraction method applied to digital collection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110266481.0A CN112990001A (en) | 2021-03-11 | 2021-03-11 | Remote sensing image monomer building rapid and accurate extraction method applied to digital collection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112990001A true CN112990001A (en) | 2021-06-18 |
Family
ID=76336390
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110266481.0A Pending CN112990001A (en) | 2021-03-11 | 2021-03-11 | Remote sensing image monomer building rapid and accurate extraction method applied to digital collection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112990001A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113610013A (en) * | 2021-08-10 | 2021-11-05 | 四川易方智慧科技有限公司 | Method for extracting building outline based on RGB (Red Green blue) wave bands of high-definition remote sensing image |
| CN114187532A (en) * | 2022-02-15 | 2022-03-15 | 中国测绘科学研究院 | Method and system for generating space-time sample of remote sensing image and intelligently iteratively classifying space-time sample |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140140571A1 (en) * | 2011-03-02 | 2014-05-22 | Raf Technology, Inc. | Document fingerprinting |
| CN108830285A (en) * | 2018-03-14 | 2018-11-16 | 江南大学 | A kind of object detection method of the reinforcement study based on Faster-RCNN |
| CN109446992A (en) * | 2018-10-30 | 2019-03-08 | 苏州中科天启遥感科技有限公司 | Remote sensing image building extracting method and system, storage medium, electronic equipment based on deep learning |
| CN110298348A (en) * | 2019-06-12 | 2019-10-01 | 苏州中科天启遥感科技有限公司 | Remote sensing image building sample areas extracting method and system, storage medium, equipment |
| CN110472544A (en) * | 2019-08-05 | 2019-11-19 | 上海英迈吉东影图像设备有限公司 | A kind of training method and system of article identification model |
| CN111460936A (en) * | 2020-03-18 | 2020-07-28 | 中国地质大学（武汉） | Remote sensing image building extraction method, system and electronic equipment based on U-Net network |
-
2021
- 2021-03-11 CN CN202110266481.0A patent/CN112990001A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140140571A1 (en) * | 2011-03-02 | 2014-05-22 | Raf Technology, Inc. | Document fingerprinting |
| CN108830285A (en) * | 2018-03-14 | 2018-11-16 | 江南大学 | A kind of object detection method of the reinforcement study based on Faster-RCNN |
| CN109446992A (en) * | 2018-10-30 | 2019-03-08 | 苏州中科天启遥感科技有限公司 | Remote sensing image building extracting method and system, storage medium, electronic equipment based on deep learning |
| CN110298348A (en) * | 2019-06-12 | 2019-10-01 | 苏州中科天启遥感科技有限公司 | Remote sensing image building sample areas extracting method and system, storage medium, equipment |
| CN110472544A (en) * | 2019-08-05 | 2019-11-19 | 上海英迈吉东影图像设备有限公司 | A kind of training method and system of article identification model |
| CN111460936A (en) * | 2020-03-18 | 2020-07-28 | 中国地质大学（武汉） | Remote sensing image building extraction method, system and electronic equipment based on U-Net network |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113610013A (en) * | 2021-08-10 | 2021-11-05 | 四川易方智慧科技有限公司 | Method for extracting building outline based on RGB (Red Green blue) wave bands of high-definition remote sensing image |
| CN114187532A (en) * | 2022-02-15 | 2022-03-15 | 中国测绘科学研究院 | Method and system for generating space-time sample of remote sensing image and intelligently iteratively classifying space-time sample |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108986119B (en) | Image segmentation method and device, computer equipment and readable storage medium | |
| CN102592268B (en) | Method for segmenting foreground image | |
| CN110298348B (en) | Method and system for extracting remote sensing image building sample region, storage medium and equipment | |
| CN112990001A (en) | Remote sensing image monomer building rapid and accurate extraction method applied to digital collection | |
| US10803355B2 (en) | Method for training image generator | |
| CN110598564B (en) | OpenStreetMap-based high-spatial-resolution remote sensing image transfer learning classification method | |
| CN113223042B (en) | Intelligent acquisition method and equipment for remote sensing image deep learning sample | |
| CN110992384B (en) | Semi-automatic image data labeling method, electronic device and storage medium | |
| CN109685821A (en) | Region growing 3D rock mass point cloud plane extracting method based on high quality voxel | |
| CN110956207B (en) | Method for detecting full-element change of optical remote sensing image | |
| US20220414892A1 (en) | High-precision semi-automatic image data labeling method, electronic apparatus, and storage medium | |
| CN111161300B (en) | Niblack image segmentation method based on improved Otsu method | |
| CN114463623A (en) | Method and device for detecting farmland change based on multi-scale remote sensing image | |
| CN115019163A (en) | City factor identification method based on multi-source big data | |
| CN111104850A (en) | Remote sensing image building automatic extraction method and system based on residual error network | |
| CN111144487B (en) | Method for establishing and updating remote sensing image sample library | |
| CN113034511A (en) | Rural building identification algorithm based on high-resolution remote sensing image and deep learning | |
| CN116822980A (en) | Base vector guided typical earth surface element intelligent monitoring method and system | |
| CN116129278A (en) | Land utilization classification and identification system based on remote sensing images | |
| CN108319927B (en) | Method for automatically identifying diseases | |
| CN113505844A (en) | Label generation method, device, equipment, storage medium and program product | |
| CN109522941B (en) | Atmospheric light intensity estimation method based on image | |
| CN117572457B (en) | Cross-scene multispectral point cloud classification method based on pseudo tag learning | |
| CN113822900B (en) | Method and system for automatically selecting new image sample based on vector constraint object-oriented | |
| CN113160262A (en) | Oyster row extraction method, system and medium based on high-resolution satellite remote sensing image |
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 |