CN107657246B - Remote sensing image building detection method based on multi-scale filtering building index - Google Patents
Remote sensing image building detection method based on multi-scale filtering building index Download PDFInfo
- Publication number
- CN107657246B CN107657246B CN201710980703.9A CN201710980703A CN107657246B CN 107657246 B CN107657246 B CN 107657246B CN 201710980703 A CN201710980703 A CN 201710980703A CN 107657246 B CN107657246 B CN 107657246B
- Authority
- CN
- China
- Prior art keywords
- building
- image
- remote sensing
- index
- scale
- 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.)
- Active
Links
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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/194—Terrestrial scenes using hyperspectral data, i.e. more or other wavelengths than RGB
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
the invention discloses a remote sensing image building detection method based on a multi-scale filtering building index. The invention is based on digital morphology and multi-scale filtering and realizes the rapid and automatic detection of the buildings of the high-resolution remote sensing images. Firstly, generating a brightness image based on the maximum value of each waveband of the multiband high-resolution remote sensing image; next, performing an on operation based on the reconstruction on the luminance image to generate an enhanced image; then, using a series of window-sized median filters for the enhanced image to obtain a multi-scale filtered differential image sequence; and averaging the sequence to obtain a multi-scale filtering building index and outputting a corresponding characteristic image. The index can effectively represent the probability of the building on the remote sensing image, and the building on the high-resolution remote sensing image can be effectively extracted based on the index.
Description
Technical Field
the invention belongs to the technical field of image processing, relates to a remote sensing image Building detection method, and particularly relates to a remote sensing image Building detection method based on a Multiscale Filtering Building Index (Multiscale Filtering Building Index).
background
Remote sensing is one of the important means of earth observation. The building is one of important ground features in the remote sensing image, and the building extraction of the remote sensing image has an important role in the aspects of city planning, general survey of geographical national conditions, GIS database updating and the like. However, in the face of high-resolution remote sensing of mass data, intelligent interpretation techniques have not been improved to a satisfactory extent. The automatic building extraction of the high-resolution remote sensing image mainly comprises the following problems:
(1) The traditional method based on the spectral information is difficult to obtain a good detection result;
the traditional remote sensing image information extraction is mostly based on spectral information, the high-resolution remote sensing image generally only has four wave bands of near infrared, red, green and blue, the number of the wave bands is obviously reduced compared with common medium-low resolution remote sensing images such as Landsat and the like, and the spectral distinguishing information of a building and other ground objects is obviously reduced; in addition, a large amount of feature detail information is provided by high spatial resolution, spectral differences in the same kind of ground objects of a building are increased, and spectral differences among different ground objects are reduced, so that the traditional spectrum-based remote sensing image information extraction method is difficult to obtain a good detection result in a high-resolution remote sensing image. In fact, buildings are easily seriously confused with objects such as roads, bare land and cement land.
(2) The complexity and the difference of the building species are very different, which causes serious difficulty for high-precision identification;
the building is easy to be seriously confused with the ground objects such as roads, bare lands, cement lands and the like on the limited wave bands of high-resolution images; the difference between the structure and the size of the building is large, buildings in rural areas and suburbs often occupy small floor area and are irregular in shape, urban buildings often occupy large floor area and are different in height, and the difference between the building shapes of gymnasiums, commercial districts and the like and residential buildings is large. Although the use of the supervised learning method is high in precision due to the above phenomena, a large amount of manpower and time are often needed for collecting training samples of different types of buildings, the generalization capability of the classification model is large, and the precision is greatly influenced by a training set; the unsupervised method has large difference of algorithm efficiency and adaptability, is difficult to find a building extraction method which is simultaneously suitable for rural areas, cities and suburbs, and has lower precision than supervised learning.
(3) The distance from intelligent extraction is far from the aspects of data volume, algorithm efficiency and precision;
At present, main geographic information production units such as a surveying and mapping bureau, a city planning bureau, a geographic information center and the like still mainly use a method for manually delineating ground objects such as buildings and the like by means of visual interpretation, so that a large amount of manpower and time are consumed, and the efficiency is low. The main reason is that the existing intelligent remote sensing interpretation technology is difficult to meet the application requirements in the aspects of data volume, time, precision and the like. Since the 21 st century, the large number of commercial high-resolution remote sensing satellites at home and abroad have risen, so that it is no longer difficult to acquire TB-level high-resolution images every day. However, as outlined in (2), the current algorithms either take a long time, require manual collection of a large number of samples, or have poor adaptability, making it difficult to extract various types of buildings efficiently; the existing storage space and computer hardware are difficult to synchronously and quickly extract buildings from mass remote sensing data, and existing high-performance equipment is difficult to popularize in basic level production units in a short period; furthermore, existing research and reference literature indicates that many building detection methods, while performing well in the experimental area of the paper, are difficult to achieve with high accuracy in other environments. Therefore, the ground features in the remote sensing image are extracted automatically and intelligently in a long distance.
In conclusion, automatic extraction of high-resolution remote sensing image buildings is a difficult problem in target identification, has rich application value and is still widely researched up to now.
disclosure of Invention
in order to solve the technical problem, the invention provides a remote sensing image building detection method based on a multi-scale filtering building index. The multi-scale filtering building index is used for representing the probability that each pixel belongs to a building, and the larger the index value is, the larger the probability that each pixel belongs to the building is.
The technical scheme adopted by the invention is as follows: a remote sensing image building detection method based on multi-scale filtering building indexes is characterized by comprising the following steps:
Step 1: constructing a multi-scale filtering building index, and specifically realizing the multi-scale filtering building index comprises the following substeps;
Step 1.1: generating a brightness image aiming at the acquired multispectral high-resolution remote sensing image;
step 1.2: enhancing the brightness image to obtain an enhanced image;
Step 1.3: acquiring a difference image sequence of multi-scale median filtering;
Step 1.4: generating a multi-scale filtering building index;
Step 2: the method comprises the following steps of automatically detecting a remote sensing image building based on a multi-scale filtering building index, and specifically realizing the method comprises the following substeps;
step 2.1: calculating a multi-scale filtering building index of each pixel aiming at the acquired multi-spectral high-resolution remote sensing image;
step 2.2: setting a threshold T according to the multi-scale filtering building index corresponding to each pixel, and judging the pixel larger than the threshold T as a building;
step 2.3: and (3) carrying out post-processing on the building image obtained in the step 2.2 to obtain a final detection result.
compared with the prior art, the invention has the following beneficial effects:
(1) The index of the building representing the high-resolution remote sensing image and the corresponding high-precision full-automatic building detection method are provided.
(2) the method can be effectively applied to automatic and rapid building extraction and special information production in surveying and mapping remote sensing and geographic information industries.
(3) the reference and reference are provided for quick, high-precision and full-automatic intelligent extraction of the thematic information of the remote sensing images in the future.
Drawings
FIG. 1 is a flow chart of building index construction with multi-scale filtering according to an embodiment of the present invention;
FIG. 2 is a flow chart of the automatic detection of a remote sensing image building according to an embodiment of the present invention;
FIG. 3 shows the experimental results and ground truth map (sample area one) of the conventional building detection method according to the embodiment of the present invention,
Wherein, (a) K mean, (b) gray level co-occurrence matrix, (c) MBI, d) MFBI, (e) ground truth map;
FIG. 4 is a diagram of the test result and ground truth (sample area two) of the common building detection method according to the embodiment of the present invention,
Wherein, (a) K mean, (b) gray level co-occurrence matrix, (c) MBI, d) MFBI, and (e) ground truth map.
Detailed Description
in order to facilitate the understanding and implementation of the present invention for those of ordinary skill in the art, the present invention is further described in detail with reference to the accompanying drawings and examples, it is to be understood that the embodiments described herein are merely illustrative and explanatory of the present invention and are not restrictive thereof.
the invention provides a remote sensing image building detection method based on a multi-scale filtering building index, which comprises the following steps:
Step 1: constructing a multi-scale filtering building index;
Referring to fig. 1, the specific implementation includes the following sub-steps;
Step 1.1: generating a brightness image aiming at the acquired multispectral high-resolution remote sensing image;
and acquiring the maximum value of each wave band of each pixel aiming at the acquired multispectral high-resolution remote sensing image to generate a brightness image.
step 1.2: performing reconstruction-based top hat operation on the brightness image to obtain an enhanced image;
Step 1.3: acquiring a difference image sequence of multi-scale median filtering;
and (3) respectively performing median filtering on the enhanced images obtained in the step (1.2) by using rectangular windows with side lengths from S min to S max and a distance delta S, and respectively performing difference on adjacent two of the obtained images to obtain a difference image sequence of multi-scale median filtering.
step 1.4: generating a multi-scale filtering building index;
And averaging all gray values corresponding to pixels at the same positions on the difference image sequence, and normalizing to be between 0 and 1 to obtain a multi-scale filtering building index (MFBI). The larger the value of each pixel after the calculation is, the higher the probability of belonging to a building is.
Step 2: automatically detecting the remote sensing image building based on the multi-scale filtering building index;
Referring to fig. 2, the specific implementation includes the following sub-steps;
step 2.1: calculating a multi-scale filtering building index of each pixel aiming at the acquired multi-spectral high-resolution remote sensing image;
This embodiment requires acquiring as many bands as possible of the image. At least three bands of near infrared, red and green are required.
Step 2.2: setting a threshold T according to the multi-scale filtering building index corresponding to each pixel, and judging the pixel larger than the threshold T as a building;
step 2.3: post-processing the building image obtained in the step 2.2 to obtain a final detection result;
For the building image obtained in the step 2.2, in the spectral information part, based on the NDVI and NDWI spectral information of each pixel, eliminating false alarms caused by vegetation water; and traversing the shape characteristic part to obtain each connected region of the binary image, performing post-processing based on geometric characteristics such as the area, the length-width ratio and the like of each connected region, eliminating false alarms, filling holes and obtaining a final detection result.
the main innovation of the invention is that:
(1) The average value of the median difference sequence in each window is used for describing the probability size of the pixel belonging to the building, so that a multi-scale filtering building index is provided.
(2) Based on the multi-scale filtering building index, an automatic building detection method is researched.
In order to verify the effectiveness of the method, two representative remote sensing image sample areas with rich building types are selected for carrying out experiments. The comparison experiment was performed using a K-means algorithm that performs classification only depending on spectral information, a gray level co-occurrence matrix algorithm that considers only spatial structure information, and a Morphological Building Index (MBI) that considers both spatial information and spectral information.
The detection results of the K-means, the gray level co-occurrence matrix, the morphological construction index (MBI), the multi-scale filtering construction index (FMBI), and the ground truth map of the sample area one and the sample area two are shown in (a) to (e) sub-graphs in fig. 3 and fig. 4, respectively. It is worth to be noted that the experiment result corresponding to the K mean value method for classifying and extracting buildings based on the spectral information is seriously confused with vegetation, roads and the like; characteristic images of the gray level co-occurrence matrix considering only image spatial information are also difficult to distinguish buildings, roads and parts of vegetation. Therefore, the two methods do not participate in the precision evaluation of building detection, and quantitative precision evaluation and analysis are only developed between the method and the Morphological Building Index (MBI). Quantitative Accuracy evaluation selects three indexes of Recall ratio (Recall), Precision ratio (Precision) and total Accuracy (Overall Accuracy); the quantitative analysis mainly takes into account the time required for both methods (including post-processing) to complete the extraction of the same sample region target.
The above experiments were all programmed and tested in a Visual Studio 2015 environment based on OpenCV 3.1.0.
Tables 1 and 2 illustrate the morphological building index and the detection accuracy of the multi-scale filtered building index presented herein; table 3 illustrates the time required for the two methods to test in two test zones.
TABLE 1 detection accuracy of MBI and MFBI in sample area one
TABLE 2 detection accuracy of MBI and MFBI in sample area two
TABLE 3 calculation time (units: seconds) for sample area one and sample area two MBI and MFBI
| MBI | MFBI | |
| Sample area one | 21.8 | 1.97 |
| Sample area two | 23.5 | 2.49 |
from the innovation points and experimental results of the method summarized above, the main advantages of the method are as follows:
(1) the multi-scale filtering building index and the corresponding building extraction method fully combine the spectral information and the spatial structure information of the image. In the process of calculating the index, the spectral information of a building is highlighted through the transformation of a brightness image and a top cap of an opening reconstruction, and the contrast with surrounding ground objects is enlarged; the edge shape information of the building is highlighted through the differential sequence image; the post-treatment process of the method further considers the geometrical characteristics of area, length-width ratio and the like. Therefore, a series of problems of traditional remote sensing image target extraction based on spectrum are fully overcome.
(2) The multi-scale filtering building index and the corresponding building extraction method have higher detection precision for buildings in different forms in cities. From the analysis in (1), the method fully considers the edge shape information of the building and the geometric characteristics such as the area, the aspect ratio and the like. In addition, through a series of detection windows with different scales, the idea is different from the traditional idea of template matching, and buildings with different shapes and sizes can be effectively detected. The idea effectively solves the problem that the detection precision is greatly influenced by the complex difference of the building species.
(3) the multi-scale filtering building index and the corresponding building extraction method have the advantages of low algorithm overhead and short calculation time. As can be seen from the program operation time counted in table 3, for an area of about one kilometer in the square circle, it takes only about 2 seconds to complete the building detection process based on the index. Meanwhile, the method has high automation degree, only needs manual parameter input, does not need any interactive processing, and reduces manpower to the maximum extent. Therefore, the method can be effectively used for quickly and automatically extracting the buildings on the large-area remote sensing image.
the above advantages can effectively solve the three difficulties of high resolution image building automatic detection proposed by the technical background part.
it should be understood that parts of the specification not set forth in detail are well within the prior art.
It should be understood that the above description of the preferred embodiments is given for clarity and not for any purpose of limitation, and that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (5)
1. A remote sensing image building detection method based on multi-scale filtering building indexes is characterized by comprising the following steps:
step 1: constructing a multi-scale filtering building index, and specifically realizing the multi-scale filtering building index comprises the following substeps;
Step 1.1: generating a brightness image aiming at the acquired multispectral high-resolution remote sensing image;
step 1.2: enhancing the brightness image to obtain an enhanced image;
step 1.3: acquiring a difference image sequence of multi-scale median filtering;
in step 1.3, median filtering is respectively performed on the enhanced image obtained in step 1.2 by using rectangular windows with side lengths from S min to S max and a distance of delta S, and differences are respectively performed on adjacent two of the obtained images to obtain a difference image sequence of multi-scale median filtering;
step 1.4: generating a multi-scale filtering building index;
step 2: the method comprises the following steps of automatically detecting a remote sensing image building based on a multi-scale filtering building index, and specifically realizing the method comprises the following substeps;
Step 2.1: calculating a multi-scale filtering building index of each pixel aiming at the acquired multi-spectral high-resolution remote sensing image;
step 2.2: setting a threshold T according to the multi-scale filtering building index corresponding to each pixel, and judging the pixel larger than the threshold T as a building;
Step 2.3: and (3) carrying out post-processing on the building image obtained in the step 2.2 to obtain a final detection result.
2. the method for detecting the buildings according to the remote sensing image based on the multi-scale filtering building index, which is characterized in that: in step 1.1, the maximum value of each wave band of each pixel is obtained, and a brightness image is generated.
3. The method for detecting the buildings according to the remote sensing image based on the multi-scale filtering building index, which is characterized in that: in step 1.2, the top hat operation based on reconstruction is carried out on the brightness image to obtain an enhanced image.
4. The method for detecting the buildings according to the remote sensing image based on the multi-scale filtering building index, which is characterized in that: in step 1.4, all gray values corresponding to pixels at the same positions on the difference image sequence are averaged and normalized to be between 0 and 1, and the multi-scale filtering building index is obtained.
5. The method for detecting the buildings according to the remote sensing image based on the multi-scale filtering building index, which is characterized in that: in step 2.3, in the spectral information part, the building image obtained in step 2.2 is subjected to false alarm removal caused by vegetation water body based on NDVI and NDWI spectral information of each pixel; and traversing the shape characteristic part to obtain each connected region of the binary image, performing post-processing based on the area and length-width ratio characteristics of each connected region, eliminating false alarms, filling holes and obtaining a final detection result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710980703.9A CN107657246B (en) | 2017-10-19 | 2017-10-19 | Remote sensing image building detection method based on multi-scale filtering building index |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710980703.9A CN107657246B (en) | 2017-10-19 | 2017-10-19 | Remote sensing image building detection method based on multi-scale filtering building index |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107657246A CN107657246A (en) | 2018-02-02 |
| CN107657246B true CN107657246B (en) | 2019-12-10 |
Family
ID=61117951
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710980703.9A Active CN107657246B (en) | 2017-10-19 | 2017-10-19 | Remote sensing image building detection method based on multi-scale filtering building index |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107657246B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111144335A (en) * | 2019-12-30 | 2020-05-12 | 自然资源部国土卫星遥感应用中心 | Method and device for building deep learning model |
| CN111323768B (en) * | 2020-02-26 | 2023-12-12 | 北京佳格天地科技有限公司 | Building change identification method, device, terminal and storage medium |
| CN111985575B (en) * | 2020-09-02 | 2022-08-02 | 四川九洲电器集团有限责任公司 | Hyperspectral image classification method based on convolutional neural network |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984947A (en) * | 2014-05-30 | 2014-08-13 | 武汉大学 | High-resolution remote sensing image house extraction method based on morphological house indexes |
| CN105608458A (en) * | 2015-10-20 | 2016-05-25 | 武汉大学 | High-resolution remote sensing image building extraction method |
-
2017
- 2017-10-19 CN CN201710980703.9A patent/CN107657246B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984947A (en) * | 2014-05-30 | 2014-08-13 | 武汉大学 | High-resolution remote sensing image house extraction method based on morphological house indexes |
| CN105608458A (en) * | 2015-10-20 | 2016-05-25 | 武汉大学 | High-resolution remote sensing image building extraction method |
Non-Patent Citations (2)
| Title |
|---|
| A Multidirectional and Multiscale Morphological Index for Automatic Building Extraction from Multispectral GeoEye-1 Imagery;Xin Huang等;《Photogrammetric Engineering & Remote Sensing》;20110731;第77卷(第7期);第721-732页 * |
| 增强型形态学建筑指数的建筑物变化检测;陶峰等;《测绘科学》;20170531;第42卷(第5期);第29-34页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107657246A (en) | 2018-02-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108596103B (en) | High-resolution remote sensing image building extraction method based on optimal spectral index selection | |
| CN111626269B (en) | Practical large-space-range landslide extraction method | |
| CN105787937B (en) | A kind of high resolution remote sensing image road change detecting method based on OSM | |
| CN111160199B (en) | Highway disaster information detection method based on high-resolution remote sensing image | |
| CN103984946B (en) | High resolution remote sensing map road extraction method based on K-means | |
| CN111062368B (en) | City update region monitoring method based on Landsat time sequence remote sensing image | |
| CN107657246B (en) | Remote sensing image building detection method based on multi-scale filtering building index | |
| CN111339948B (en) | Automatic identification method for newly-added buildings of high-resolution remote sensing images | |
| CN107688777B (en) | Urban green land extraction method for collaborative multi-source remote sensing image | |
| CN110569751B (en) | High-resolution remote sensing image building extraction method | |
| CN110309781A (en) | Damage remote sensing recognition method in house based on the fusion of multi-scale spectrum texture self-adaption | |
| CN113657324A (en) | Urban functional area identification method based on remote sensing image ground object classification | |
| CN106971397B (en) | Based on the city high-resolution remote sensing image dividing method for improving JSEG algorithms | |
| CN104463836A (en) | City green space remote-sensing measuring method based on moving windows | |
| CN106650663B (en) | The determination method of building true and false variation and pseudo- containing the method change minimizing technology | |
| CN114241321A (en) | Rapid and accurate identification method for high-resolution remote sensing image flat-topped building | |
| Chehata et al. | Object-based forest change detection using high resolution satellite images | |
| CN113902759B (en) | Space-spectrum information combined satellite-borne hyperspectral image segmentation and clustering method | |
| CN104966091A (en) | Strip mine road extraction method based on unmanned plane remote sensing images | |
| CN113096114B (en) | High-resolution urban water body pattern spot remote sensing extraction method combining morphology and index | |
| CN106324708B (en) | Digitizing solution, the device of rainfall record drawing | |
| Ding et al. | Study on building extraction from high-resolution images using Mbi | |
| Manandhar et al. | Segmentation based building detection in high resolution satellite images | |
| CN111340761B (en) | Remote sensing image change detection method based on fractal attribute and decision fusion | |
| CN110929739B (en) | Automatic impervious surface range remote sensing iterative extraction method |
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 |