Color steel tile building extraction method and system based on PlanetScope image

Abstract

The invention discloses a color steel tile building extraction method and system based on a PlanetScope image, and relates to the technical field of remote sensing image processing, and the method comprises the steps: obtaining image data, and carrying out the coordinate system conversion and marking of the image data, and obtaining a binary image; performing spectral feature extraction on the binary image to obtain spectral features; performing amplification operation on the spectral features, constructing a feature set, and inputting the feature set into a random forest model to obtain feature parameters; and constructing a feature combination based on the feature parameters, and performing threshold search on the feature combination to realize extraction of the color steel tile building. According to the method, the genetic algorithm is utilized to perform feature combination optimization and threshold search, so that the color steel tile building identification efficiency can be greatly improved, and color steel tile buildings in different areas and complex environments can be processed; by searching the optimal threshold value of the feature combination, efficient recognition of the color steel tile building is ensured, and the applicability and operability of remote sensing image analysis in practical application are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing image processing, and particularly to a method and system for extracting color steel tile buildings based on PlanetScope images. Background Art

[0002] In remote sensing image analysis, accurately extracting color steel tile buildings is not only of great significance for urban planning and building statistics, but also plays a key role in the monitoring and management of potential illegal buildings; due to their material and spectral reflection characteristics, color steel tile buildings are easily recognizable in high-resolution images; however, in a complex urban background, the dense building distribution, the similarity of different building materials, and the interference of vegetation often pose challenges to the recognition process, and these problems are particularly prominent in the detection of illegal buildings because illegal buildings usually have a chaotic layout, diverse materials, and are easily mixed with the surrounding environment. Therefore, developing a remote sensing analysis method that can accurately extract color steel tile buildings can not only effectively identify regular buildings, but also provide important technical support for the rapid positioning and management of potential illegal buildings.

[0003] However, traditional building extraction methods often rely on satellite images with relatively low resolution, such as Landsat or Sentinel-2 images. Although these images have a wide coverage range, their spatial resolution is relatively low. The resolution of Sentinel-2 images is about 10 meters, which limits the extraction of small buildings, especially color steel tile buildings, and it is impossible to clearly identify the detailed structure of the buildings; therefore, the accuracy and reliability of color steel tile building extraction are relatively limited, especially in a complex urban environment; in contrast, PlanetScope images have a spatial resolution of up to about 3 meters, providing more detailed spatial information support for color steel tile building extraction; the higher resolution can effectively capture the boundary features and material reflection differences of color steel tile buildings, further improving the recognition accuracy of buildings; in addition, the spectral information of PlanetScope images can be combined with advanced machine learning algorithms, such as random forest or genetic algorithm, to optimize the feature combination and improve the classification accuracy; this extraction method based on high-resolution satellite images not only has higher accuracy in building recognition, but also provides important technical support for future large-scale urban building monitoring. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is how to provide a method for extracting color steel tile buildings based on PlanetScope images, which reduces the problems of over-reliance on manual modification and limited resolution acquisition existing in the prior art.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a method for extracting color steel tile buildings based on PlanetScope images, which includes obtaining image data, performing coordinate system conversion and annotation on the image data to obtain a binary image; extracting spectral features from the binary image to obtain spectral features; performing an amplification operation on the spectral features, constructing a feature set and inputting the feature set into a random forest model to obtain feature parameters; constructing a feature combination based on the feature parameters, and performing threshold search on the feature combination to achieve the extraction of color steel tile buildings.

[0008] As a preferred solution of the method for extracting color steel tile buildings based on PlanetScope images of the present invention, wherein: obtaining the feature parameters includes the following steps: performing an amplification operation on the spectral features to obtain an amplified feature set; inputting the amplified feature set into a random forest model and training it, and calculating the importance values of each feature; traversing each node of the random forest model to obtain the frequency, maximum value, and minimum value of each feature appearing in the random forest model.

[0009] As a preferred solution of the method for extracting color steel tile buildings based on PlanetScope images of the present invention, wherein: constructing a feature combination based on the feature parameters means that after comparing the importance values and frequencies of each feature, using the frequency as a weight for feature sampling without replacement to construct a feature combination.

[0010] As a preferred solution of the method for extracting color steel tile buildings based on PlanetScope images of the present invention, wherein: performing threshold search on the feature combination includes the following steps: inputting the binary image, the maximum value of each feature, the minimum value of each feature, and the feature combination into a GA algorithm constructed based on a Python environment; evaluating the classification performance of different feature combinations by using the fitness function of the GA algorithm to obtain the feature combination with the highest classification performance.

[0011] As a preferred solution of the method for extracting color steel tile buildings based on PlanetScope images of the present invention, wherein: obtaining the image data means that since the study area contains color steel tile buildings of different areas, obtaining the image data in the study area through the official website of the PlanetScope satellite.

[0012] As a preferred solution of the method for extracting color steel tile buildings based on PlanetScope images according to the present invention, wherein: the coordinate system conversion refers to converting the coordinate system of the image data from WGS_1984_UTM_Zone_49N to GCS_WGS_1984; the annotation refers to using ArcMap software to perform pixel annotation on the image data after coordinate system conversion for color steel tile buildings and non-color steel tile buildings, so as to obtain a binary image containing color steel tile building pixels and non-color steel tile building pixels.

[0013] As a preferred solution of the method for extracting color steel tile buildings based on PlanetScope images according to the present invention, wherein: the spectral feature extraction refers to using the value extraction to point tool supported by ArcMap software to perform spectral feature extraction on the binary image, so as to obtain the spectral features of color steel tile building and non-color steel tile building samples.

[0014] In a second aspect, to further solve the security problems existing in remote sensing image processing, the embodiments of the present invention provide a system for extracting color steel tile buildings based on PlanetScope images, which includes: an image acquisition module for acquiring image data in the study area and performing coordinate system conversion and annotation to obtain a binary image containing color steel tile building pixels and non-color steel tile building pixels; a feature extraction module for using the value extraction to point tool to perform spectral feature extraction on the binary image to obtain the spectral features of color steel tile building and non-color steel tile building samples; a feature parameter module for performing amplification operation on the spectral features, constructing a feature set and inputting the feature set into a random forest model to obtain feature parameters; a threshold search module for constructing a feature combination based on the feature parameters and performing optimal threshold search on the feature combination to achieve the extraction of color steel tile buildings.

[0015] In a third aspect, the embodiments of the present invention provide a computer device, including a memory and a processor, wherein: when the computer program is executed by the processor, it implements any step of the method for extracting color steel tile buildings based on PlanetScope images as described in the first aspect of the present invention.

[0016] In a fourth aspect, the embodiments of the present invention provide a computer-readable storage medium, on which a computer program is stored, wherein: when the computer program is executed by the processor, it implements any step of the method for extracting color steel tile buildings based on PlanetScope images as described in the first aspect of the present invention.

[0017] Advantages of the present invention: The present invention uses a genetic algorithm for feature combination optimization and threshold search to construct a method for extracting color steel tile buildings based on PlanetScope images, which can greatly improve the recognition efficiency of color steel tile buildings, process color steel tile buildings in different regions and complex environments, and contribute to improving data reliability and analysis accuracy in urban planning, environmental monitoring and related applications; by searching for the optimal threshold of feature combinations, efficient recognition of color steel tile buildings is ensured, and the applicability and operability of remote sensing image analysis in practical applications are enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings. Among them:

[0019] Figure 1 It is the overall flowchart of the method for extracting color steel tile buildings based on PlanetScope images in Embodiment 1.

[0020] Figure 2 It is the structural schematic diagram of the computer device in Embodiment 3.

[0021] Figure 3 It is the feature importance and frequency diagram of the random forest model trained by the present invention in Embodiment 4.

[0022] Figure 4 It is the schematic diagram of the color steel tile buildings finally extracted in the study area in Embodiment 4.

[0023] Figure 5 It is the feature comparison box diagram of the color steel tile building classification rules constructed by the present invention in Embodiment 4. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] In order to make the above objects, features and advantages of the present invention more obvious and understandable, the following will give a detailed description of the specific embodiments of the present invention in conjunction with the drawings of the specification.

[0025] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0026] Second, the "one embodiment" or "embodiment" referred to herein means a specific feature, structure or characteristic that may be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not all refer to the same embodiment, nor is it an individual or selectively mutually exclusive embodiment with other embodiments.

[0027] Embodiment 1

[0028] Referring to Figure 1 , which is the first embodiment of the present invention. This embodiment provides a method for extracting color steel tile buildings based on PlanetScope images.

[0029] The existing building extraction methods mainly have the following problems: Traditional building extraction methods often rely on satellite images with relatively low resolution, such as Landsat or Sentinel-2 images. Although such images have a wide coverage range, their spatial resolution is relatively low. The resolution of Sentinel-2 images is about 10 meters, which limits the extraction of small buildings, especially color steel tile buildings, and it is impossible to clearly identify the detailed structure of the buildings; Therefore, the accuracy and reliability of color steel tile building extraction are relatively limited, especially in complex urban environments.

[0030] This application can effectively solve the above-mentioned problems. Next, multiple embodiments will be combined to elaborate in detail how to implement the method for extracting color steel tile buildings based on PlanetScope images.

[0031] Figure 1 Fig. shows the overall flowchart of the method for extracting color steel tile buildings based on PlanetScope images, including:

[0032] S1: Obtain image data, perform coordinate system conversion and annotation on the image data, and obtain a binary image.

[0033] Preferably, obtaining image data means that since the study area contains color steel tile buildings with different areas, the PSB.SD image data in the study area, that is, Super Dove image data, is obtained through the PlanetScope satellite official website, and the format of the image data is.tif.

[0034] Furthermore, the coordinate system conversion means converting the coordinate system of the PSB.SD image data from WGS_1984_UTM_Zone_49N to GCS_WGS_1984.

[0035] Further, annotation refers to using ArcMap software to perform pixel annotation on the PSB.SD image data after coordinate system conversion for color steel tile buildings and non-color steel tile buildings, thereby obtaining a binary image containing color steel tile building pixels and non-color steel tile building pixels. The annotation result is an 8-bit binary image, where the color steel tile building pixel value is 0 and the non-color steel tile building pixel value is 255.

[0036] Preferably, the selected PlanetScope's SuperDove image data of the present invention has unique advantages. The resolution of the Super Dove sensor is particularly suitable for identifying medium-scale color steel tile buildings, which can balance computational efficiency while ensuring accuracy; by choosing to convert the coordinate system of the image data from UTM projection to GCS_WGS_1984 geographic coordinate system, not only the consistency of spatial reference is ensured, but more importantly, it facilitates subsequent overlay analysis with other geographic information data, avoids deformation problems at the junction of different regional projection zones, and improves the accuracy of large-scale regional analysis.

[0037] S2: Extract spectral features from the binary image to obtain spectral features.

[0038] Preferably, spectral feature extraction refers to using the Extract Values to Points tool supported by ArcMap software to extract spectral features from the binary image, obtaining the spectral features of color steel tile building and non-color steel tile building samples, where the output file is in.csv format.

[0039] Preferably, the present invention uses the Extract Values to Points tool for spectral feature extraction, which can not only retain the spatial position information of pixels but also obtain spectral features, solving the problem that spatial relationships are often ignored in traditional methods.

[0040] S3: Perform an amplification operation on the spectral features, construct a feature set, and input the feature set into a random forest model to obtain feature parameters.

[0041] Preferably, obtaining the feature parameters includes the following steps: performing an amplification operation on the spectral features to obtain an amplified feature set.

[0042] Input the amplified feature set into the random forest model and train it to calculate the importance values of each feature.

[0043] Traverse each node of the random forest model to obtain the frequency, maximum value, and minimum value of each feature appearing in the random forest model.

[0044] Preferably, through the feature amplification operation, a double screening is achieved based on the random forest model. Not only the discrimination ability of features is evaluated using the importance value, but also the stability of the evaluation features is counted through frequency. This double screening mechanism significantly improves the reliability of feature selection and the generalization ability of the model.

[0045] S4: Construct a feature combination based on the feature parameters, and perform a threshold search on the feature combination to achieve the extraction of color steel tile buildings.

[0046] Preferably, constructing a feature combination based on the feature parameters means that after comparing the importance values and frequencies of each feature, sampling features without replacement with the frequency as the weight to construct the feature combination, where the number of output feature combinations is 1000.

[0047] Furthermore, performing a threshold search on the feature combination includes the following steps: input the binary image, the maximum value of each feature, the minimum value of each feature, and the feature combination into the GA algorithm constructed based on the Python environment to search for the optimal threshold of different feature combinations.

[0048] The classification performance of different feature combinations is evaluated through the fitness function of the GA algorithm, and the feature combination with the highest classification performance is the method for extracting color steel tile buildings constructed in the present invention.

[0049] It should be noted that for features with high importance, their frequencies do not necessarily show high values, while features with high frequencies mean that they are selected multiple times in the model for the classification process, indicating that the feature has strong generalization ability in the entire classification task and will not fluctuate greatly due to different data. Therefore, the frequency is selected as the feature weight.

[0050] Specifically, searching for the optimal threshold of different feature combinations means using the Genetic Algorithm (GA) to perform the threshold search of the feature combination. The algorithm runs in the Python environment, and the input data is an image file in.tif format. The genetic algorithm uses the Intersection over Union (IOU) as the fitness function to accelerate the convergence process of the model. The calculation formula of the intersection over union IOU is as follows:

[0051]

[0052] Among them, IOU is the fitness function; A is the predicted shadow classification result; B is the shadow annotation result of the pixel annotation of the color steel tile building and the non-color steel tile building.

[0053] Preferably, by proposing to use the frequency of feature appearance as the weight for feature sampling, the present invention can better reflect the stability and generalization ability of the feature, effectively reducing the over-reliance of the model on a single feature; the IOU index takes into account both the accuracy and integrity of the classification, can accelerate the model convergence, reduce the consumption of computing resources, and at the same time, since the color steel tile building often presents a regular geometric shape, using IOU as the fitness function is more suitable for the characteristics of the color steel tile building.

[0054] In summary, the present invention uses a genetic algorithm for feature combination optimization and threshold search, and constructs a method for extracting color steel tile buildings based on PlanetScope images, which can greatly improve the recognition efficiency of color steel tile buildings, process color steel tile buildings in different regions and complex environments, and contribute to improving the data reliability and analysis accuracy in urban planning, environmental monitoring and related applications; by searching for the optimal threshold of the feature combination, the efficient recognition of color steel tile buildings is ensured, and the applicability and operability of remote sensing image analysis in practical applications are enhanced.

[0055] Embodiment 2 is an embodiment of the present invention, which provides a system for extracting color steel tile buildings based on PlanetScope images, including: an image acquisition module for acquiring image data in the study area, performing coordinate system conversion and annotation, and obtaining a binary image containing color steel tile building pixels and non-color steel tile building pixels; a feature extraction module for using the value extraction to point tool to perform spectral feature extraction on the binary image to obtain the spectral features of color steel tile building and non-color steel tile building samples; a feature parameter module for performing amplification operations on the spectral features, constructing a feature set and inputting the feature set into a random forest model to obtain feature parameters; a threshold search module for constructing a feature combination based on the feature parameters and performing optimal threshold search on the feature combination to achieve the extraction of color steel tile buildings.

[0056] Embodiment 3 is an embodiment of the present invention, which is different from the previous embodiment in that:

[0057] As Figure 2 shown, if the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0058] The logic and / or steps represented in the flowchart or otherwise described herein can, for example, be considered as a definitional sequence of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.

[0059] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection portion having one or more wirings (electronic device), a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, as the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then storing it in a computer memory.

[0060] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0061] Example 4, an embodiment of the present invention, provides a method for extracting color steel tile buildings based on PlanetScope images. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0062] In this example, based on the requirement of a certain research area to contain color steel tile buildings of different sizes, PlanetScope satellite imagery was selected as the data source. The imagery of the research area obtained was the cloud-free PlanetScope imagery on October 11, 2024, and it was cropped based on the boundary. The obtained imagery dataset had completed image preprocessing, including radiometric calibration and atmospheric correction, and the data obtained only contained four image bands, namely blue (B2), green (B3), red (B4), and near-infrared (NIR), and the image format was.tif.

[0063] Secondly, a coordinate system conversion operation was performed on the obtained research area imagery data, that is, the research area imagery data was converted from the WGS_1984_UTM_Zone_49N format to the CGS_WGS_1984 format. The imagery after coordinate conversion would facilitate subsequent data operations; meanwhile, the annotation work of color steel tile buildings and non-color steel tile buildings would be based on this imagery, and the software used was ArcMap. The annotation result was an 8-bit binary image, where color steel tile buildings were represented by 0 and non-color steel tile buildings were represented by 255.

[0064] In addition, based on the binary image after annotation processing, color steel tile buildings and non-color steel tile building samples were obtained using the ArcMap software, and the spectral features of each sample were obtained using the value extraction to point tool supported by the ArcMap software. The output feature set format was.csv.

[0065] Subsequently, in the Python environment, feature augmentation was performed on the obtained spectral features of each sample. The augmentation operations included "+", "-", and " / " operations. The specific features included sum features (B2 + B3, B2 + B4, B2 + B5, B3 + B4, B3 + B5, and B4 + B5), subtraction features (B2 + B3, B2 - B4, B2 - B5, B3 - B4, B3 - B5, and B4 - B5), ratio features (B2 / B3, B2 / B4, B2 / B5, B3 / B2, B3 / B4, B3 / B5, B4 / B2, B4 / B3, B4 / B5, B5 / B2, B5 / B3, and B5 / B4), and other features ((B2 + B3) / B2, (B2 + B3) / B3, (B2 + B5) / B5, (B3 + B4) / B2, (B3 + B5) / B2, (B4 + B5) / B2, (B2 - B3) / B2, (B2 - B3) / B3, (B2 - B4) / B2, (B2 + B4) / B2, (B3 - B5) / B2, (B4 - B5) / B2, (B4 - B5) / B5, (B4 - B5) / B4, and (B4 - B5) / B3), a total of 43 features; the constructed feature set was input into the random forest model provided by the Python environment, and the importance values of different features were output as Figure 3As shown, the top 10 important features are presented; meanwhile, by traversing each node of the random forest, the frequencies, maximum values, and minimum values of different features appearing in the model are finally obtained. The information on the frequencies of appearance is as Figure 3 shown, which facilitates subsequent feature combination and threshold search. Among the parameters of the random forest model, except that the number of trees is set to 500, the remaining parameters remain default.

[0066] After comparing the importance values of each feature with the frequency information, the present invention selects the feature frequency information as the feature weight for feature combination, with the number being 1000. The number of feature combinations is strictly limited to within 3 to prevent the emergence of complex feature combinations; finally, the present invention uses the Genetic Algorithm (GA) to search for the optimal threshold of different feature combinations. The parameters of GA are set as follows: the population size is 200, the maximum number of iterations is 200 times, the upper and lower bounds of the search space have been defined in the amplification operation step, the fitness function is the Intersection over Union (IOU), and the remaining parameters remain default.

[0067] By analyzing the fitness values, a rule-based color steel tile building classification method is finally constructed, as Figure 4 shown. The color steel tile building extraction result achieved based on the feature combination constructed according to the present invention is presented. The finally obtained feature combination is "B3 / B2 < 1.0848 & (B2 + B5) / B5 > 1.4676", Figure 5 and the box plot comparing the classification features of color steel tile buildings is shown. This box plot effectively shows the distribution of different category samples in the feature space, can intuitively point out the separability of features, and by comparing the feature distributions of color steel tile buildings and non-color steel tile buildings, it helps to understand the constructed classification rules, indicating that the present invention ensures the efficient recognition of color steel tile buildings and improves the recognition efficiency of color steel tile buildings.

[0068] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for extracting colored steel tile buildings based on PlanetScope images, characterized in that: include: Acquire image data and perform coordinate system conversion and annotation on the image data to obtain a binary image; Extracting spectral features from the binary image to obtain spectral features; Performing an amplification operation on the spectral features, constructing a feature set and inputting the feature set into a random forest model to obtain feature parameters; A feature combination is constructed based on the feature parameters, and a threshold search is performed on the feature combination to achieve the extraction of color steel tile buildings.

2. The method for extracting colored steel tile buildings based on PlanetScope images as claimed in claim 1, characterized in that: The obtaining of characteristic parameters comprises the following steps: Performing an amplification operation on the spectral features to obtain an amplified feature set; The amplified feature set is input into the random forest model and trained to calculate the importance value of each feature; Each node of the random forest model is traversed to obtain the frequency, maximum value and minimum value of each feature in the random forest model.

3. The method for extracting colored steel tile buildings based on PlanetScope images as claimed in claim 2, characterized in that: Constructing a feature combination based on the feature parameters means comparing the importance value and frequency of each feature, and then using the frequency as a weight to perform feature sampling without replacement to construct a feature combination.

4. The method for extracting colored steel tile buildings based on PlanetScope images as claimed in claim 3, characterized in that: Performing a threshold search on the feature combination comprises the following steps: Input the binary image, the maximum value of each feature, the minimum value of each feature, and the feature combination into a GA algorithm built based on a Python environment; By using the fitness function of the GA algorithm to evaluate the classification performance of different feature combinations, the feature combination with the highest classification performance is obtained.

5. The method for extracting colored steel tile buildings based on PlanetScope images as claimed in claim 4, characterized in that: The acquisition of image data refers to the acquisition of image data in the study area by using the PlanetScope satellite official website because the study area contains colored steel tile buildings of different areas.

6. The method for extracting colored steel tile buildings based on PlanetScope images as claimed in claim 5, characterized in that: The coordinate system conversion refers to converting the coordinate system of the image data from WGS_1984_UTM_Zone_49N to GCS_WGS_1984; The labeling refers to labeling the pixels of the colored steel tile buildings and the non-colored steel tile buildings on the image data after the coordinate system conversion using ArcMap software, thereby obtaining a binary image containing the pixels of the colored steel tile buildings and the pixels of the non-colored steel tile buildings.

7. The method for extracting colored steel tile buildings based on PlanetScope images as claimed in claim 6, characterized in that: The spectral feature extraction refers to extracting spectral features from the binary image using the value extraction to point tool supported by ArcMap software to obtain spectral features of color steel tile building and non-color steel tile building samples.

8. A system for extracting colored steel tile buildings based on PlanetScope images, based on the method for extracting colored steel tile buildings based on PlanetScope images according to any one of claims 1 to 7, characterized in that: include, An image acquisition module is used to acquire image data in the study area and perform coordinate system conversion and annotation to obtain a binary image containing color steel tile building pixels and non-color steel tile building pixels; The feature extraction module is used to extract spectral features from binary images using the value extraction to point tool to obtain spectral features of samples of color steel tile buildings and non-color steel tile buildings; The feature parameter module is used to amplify the spectral features, construct a feature set and input the feature set into the random forest model to obtain feature parameters; The threshold search module is used to construct feature combinations based on feature parameters and perform optimal threshold search on the feature combinations to achieve the extraction of color steel tile buildings.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for extracting colored steel tile buildings based on PlanetScope images described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for extracting colored steel tile buildings based on PlanetScope images described in any one of claims 1 to 7 are implemented.

Images