An automated quality inspection method for a large amount of multi-source satellite remote sensing images

Abstract

The present invention provides an automated quality inspection method for a large amount of multi-source satellite remote sensing images, which relates to the field of remote sensing image quality inspection. The method includes the following steps: obtaining multi-source satellite remote sensing images, then performing a specification inspection on the remote sensing images to obtain a specification inspection result; then calculating the cloud amount of the remote sensing images after the specification inspection, and sampling the cloud amount image information according to preset conditions based on the cloud amount calculation result to obtain a sampling database. Then, continue to perform geometric quality inspection and radiation quality inspection on the cloud amount image information in the sampling database. Finally, comprehensively utilize the specification inspection result, cloud amount calculation result, geometric quality inspection result, and radiation quality inspection result obtained in the above steps to obtain a final inspection result. The entire method process is simple and efficient, can meet the quality inspection tasks of daily large amounts of data, and improve the quality inspection efficiency and quality.

Description

Technical Field

[0001] The present invention relates to the field of remote sensing image quality inspection, and more particularly, to an automated quality inspection method for massive multi-source satellite remote sensing images. Background Art

[0002] With the continuous launch of remote sensing satellites, the amount of remote sensing image data is increasing day by day. The daily acquired remote sensing image data across the country has reached the PB level. During the processes of satellite on-board devices, satellite attitude and orbit, remote sensing image transmission, ground reception and processing, production preprocessing, decompression, and bit reduction, etc., image data quality problems will inevitably occur, and the traditional method mainly based on interactive quality inspection can no longer meet the quality inspection tasks of daily massive data.

[0003] With the continuous improvement of high spatial resolution remote sensing image and hyperspectral remote sensing image technologies, compared with traditional satellite remote sensing images, current remote sensing satellite images have higher spatial resolution, richer spatial texture, geometric information, and spectral information, and at the same time, the satellite types are more diversified. The diversified satellites also bring different situations in data structure, spectral information characteristics, metadata information, and structure, etc. However, the information feature extraction and quality inspection methods adopted in the prior art cannot effectively control the quality of massive multi-source satellite data products. Summary of the Invention

[0004] The purpose of the present invention is to provide an automated quality inspection method for massive multi-source satellite remote sensing images, which can meet the quality inspection tasks of daily massive data and improve the quality inspection efficiency and quality.

[0005] The embodiments of the present invention are implemented as follows:

[0006] In a first aspect, an embodiment of the present application provides an automated quality inspection method for massive multi-source satellite remote sensing images, which includes the following steps:

[0007] Obtain multi-source satellite remote sensing images to obtain original image information;

[0008] Conduct a specification inspection on the original image information to obtain a specification inspection result;

[0009] Conduct cloud amount calculation on the original image information after the specification inspection to obtain a cloud amount calculation result and cloud amount image information including cloud amount percentage information;

[0010] Sample the cloud amount image information according to preset conditions to obtain a sampling database;

[0011] Conduct geometric quality inspection on the cloud amount image information in the sampling database to obtain a geometric quality inspection result and geometric quality inspection image information;

[0012] Perform radiation quality inspection on the geometric quality inspection image information to obtain the radiation quality inspection result;

[0013] Use the specification inspection result, cloud amount calculation result, geometric quality inspection result, and radiation quality inspection result to obtain the final inspection result.

[0014] In some embodiments of the present invention, the above-mentioned specification inspection of the original image information specifically includes at least one of metadata quality detection, file integrity detection, data consistency detection, data standardization detection, or data validity detection.

[0015] In some embodiments of the present invention, the above-mentioned step of performing geometric quality inspection on the cloud amount image information in the sampling database to obtain the geometric quality inspection result and geometric quality inspection image information specifically includes:

[0016] Perform similarity encoding on the cloud amount image information in the sampling database and the preset reference image information respectively to obtain the cloud amount image code and the reference image code;

[0017] Use the distance calculation method to calculate the similarity distance between the cloud amount image code and the reference image code;

[0018] Compare the similarity distance with the preset similarity threshold, and generate a geometric quality detection image according to the obtained judgment result.

[0019] In some embodiments of the present invention, the above-mentioned radiation quality inspection includes at least one of inter-slice radiation anomaly detection, inter-slice color difference anomaly detection, color cast detection, missing detection, garbled code detection, or stripe noise detection.

[0020] In some embodiments of the present invention, the above-mentioned inter-slice radiation anomaly detection specifically includes:

[0021] Calculate the gradient values of the projections of all image layers in the geometric quality inspection image information respectively to obtain the image layer gradient value information of the corresponding layers;

[0022] Obtain the first gradient maximum value and the first gradient mean value in the image layer gradient value information of the corresponding layers respectively, and record the ratio of the first gradient maximum value to the first gradient mean value as r1;

[0023] Perform edge detection on the projections of all image layers in the geometric quality inspection image information respectively to obtain the edge detection images of the corresponding layers;

[0024] Obtain the second gradient maximum value and the second gradient mean value in the edge detection images of the corresponding layers respectively, and record the ratio of the second gradient maximum value to the second gradient mean value as r2;

[0025] Judge the size relationship between r1 and r2 of the corresponding layer respectively. If r1 ≥ r2, it is recorded that there is no inter-chip radiation anomaly problem. If r1 < r2, it is recorded that there is an inter-chip radiation anomaly problem.

[0026] In some embodiments of the present invention, the above color cast detection is to automatically detect the color cast image under CIE Lab.

[0027] In some embodiments of the present invention, the above missing detection specifically includes:

[0028] Obtain the double-peak histogram image information of all frames in the geometric quality inspection image information;

[0029] Use the adaptive threshold method to segment the double-peak histogram image information. If it is lower than the adaptive threshold, the block area is marked as the missing area;

[0030] Use the fast non-local means denoising method to perform edge detection on the missing area to obtain the boundary information of the missing area;

[0031] Generate missing detection result information using the boundary information of the missing area.

[0032] In a second aspect, an embodiment of the present application provides an electronic device, which includes a memory for storing one or more programs; a processor. When the above one or more programs are executed by the above processor, the method described in any one of the above first aspects is implemented.

[0033] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in any one of the above first aspects is implemented.

[0034] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0035] By deeply analyzing the data characteristics, inspecting the data specifications, cloud amount, geometric accuracy, radiation statistical characteristic values, radiation anomalies and other indicators of the massive multi-source satellite remote sensing image products, a comprehensive evaluation can be carried out on them, and a comprehensive evaluation of the remote sensing image products under multiple indicators can be realized, thereby improving the accuracy and reliability of the detection. In addition, by sampling the cloud amount image information including the cloud amount ratio information, a large amount of data can be processed with a smaller computing memory, improving the data processing efficiency. The whole process is simple and efficient, and can effectively detect and control the quality of the massive multi-source satellite data products, which has important significance for the quality evaluation of remote sensing images. Description of the Drawings

[0036] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0037] Figure 1 It is a flowchart of an embodiment of an automated quality inspection method for a large amount of multi-source satellite remote sensing images of the present invention;

[0038] Figure 2 It is a specific flowchart of the steps for geometric quality inspection of cloud amount image information in the sampling database in the embodiments of the present invention to obtain geometric quality inspection results and geometric quality inspection image information;

[0039] Figure 3 It is a specific flowchart of inter-slice radiation anomaly detection in the embodiments of the present invention;

[0040] Figure 4 It is a specific flowchart of missing detection in the embodiments of the present invention;

[0041] Figure 5 It is a structural block diagram of an electronic device provided in the embodiments of the present invention.

[0042] Icons: 101, memory; 102, processor; 103, communication interface. Specific Embodiments

[0043] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Usually, the components of the embodiments of the present application described and illustrated in the drawings here can be arranged and designed in various different configurations.

[0044] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the present application claimed, but merely represents the selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present application.

[0045] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0046] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0047] Example

[0048] See also Figure 1 The method for automatically checking the quality of massive multi-source satellite remote sensing images comprises the following steps:

[0049] Step S1: Acquire multi-source satellite remote sensing images to obtain original image information.

[0050] In the above steps, by acquiring multi-source satellite remote sensing images, original data support is provided for subsequent quality inspection. Exemplarily, the multi-source satellite remote sensing images can be acquired in real time through corresponding satellites, or by autonomously uploading the acquired data sources.

[0051] Step S2: Perform a specification inspection on the original image information to obtain a specification inspection result.

[0052] See also Figure 1 The above-mentioned specification inspection of the original image information specifically includes: at least one of metadata quality inspection, file integrity inspection, data consistency inspection, data standardization inspection or data validity inspection.

[0053] In the above steps, by performing a specification check on the original image information, it is possible to detect whether the original image information obtained has problems in the basic file quality aspects such as file integrity, validity, and metadata quality. If there are problems, the subsequent quality inspection process can be stopped in time, or the subsequent quality inspection process can be continued after remediation, which can effectively save the system's computing resources. Among them, the file integrity check can include checking the number of data files, the number of image bands, and the integrity of the corresponding files, the data validity check can include checking the format standardization and availability of various types of file data, and the metadata quality check can include checking the integrity of metadata items and the validity of the value range.

[0054] Exemplarily, the evaluation criteria for the specification inspection may include the following four aspects: integrity, consistency, accuracy, and timeliness. To evaluate whether the original image information meets the expected specification quality requirements, it can be judged through these four aspects. Among them, integrity refers to whether there is a situation of missing data information in the original image information data. The situation of missing original image information data may be the missing record of the entire original image information data, or the missing record of a certain field information in the original image information data. The value of incomplete original image information data will be greatly reduced, and it is also the most basic evaluation criterion for the quality of original image information data. Consistency means whether the original image information data follows a unified specification, and whether the original image information data set maintains a unified format. The consistency of the quality of original image information data is mainly reflected in the specification of the original image information data record and whether the original image information data is logical. Specification means that an original image information data has its specific format. Accuracy refers to whether there are abnormalities or errors in the information recorded in the original image information data. The original image information data with accuracy problems is not only inconsistent in rules. The most common accuracy error in the original image information data is garbled characters. Secondly, the original image information data that is abnormally large or small is also the original image information data that does not meet the conditions. The accuracy of the quality of the original image information data may exist in individual records or in the entire original image information data set, such as an error in the order of magnitude record. Such errors can be audited using the statistics of the maximum and minimum values. Generally, the original image information data conforms to the law of normal distribution. If some original image information data with a small proportion has problems, then a judgment can be made by comparing the proportion of other original image information data with a small quantity. Timeliness refers to the time interval from the generation of the original image information data to the time when it can be viewed, also called the latency duration of the original image information data. Timeliness does not have high requirements for the analysis of the original image information data itself, but if the analysis period of the original image information data plus the time for establishing the original image information data is too long, it may lead to the conclusion drawn from the analysis losing its reference significance.

[0055] Step S3: Perform cloud amount calculation on the original image information after the specification inspection to obtain the cloud amount calculation result and the cloud amount image information including the cloud amount ratio information.

[0056] In the above steps, the presence of clouds, fog, shadows, etc. will cause the ground objects to be unrecognizable. That is to say, when there are a large number of clouds in the remote sensing image, it will affect the quality of the remote sensing image and seriously reduce the utilization rate of the remote sensing image. Therefore, by performing cloud amount calculation on the original image information after the specification inspection, the cloud amount ratio information can be clearly understood, thus facilitating the quality analysis of it.

[0057] Exemplarily, the original image information in the sampling database can be used to calculate cloud amount through deep learning, so as to efficiently and automatically obtain the cloud amount ratio information. And a cloud mask can be generated by deep learning and saved for subsequent applications.

[0058] Step S4: Sample the cloud amount image information according to preset conditions to obtain a sampling database.

[0059] In the above steps, the cloud amount image information can be sampled according to actual needs, so as to collect part or all of the data as samples for data analysis. Thus, a large amount of data can be processed with less computing memory, improving the data processing efficiency.

[0060] Exemplarily, sampling methods such as simple random sampling, stratified sampling, systematic sampling, and cluster sampling can be selected according to the actual system memory situation. Specifically, band extraction and sample interactive selection extraction can be performed on the cloud amount image information data. Additionally, the preset conditions can be different extractions according to the cloud amount ratio.

[0061] Step S5: Perform geometric quality inspection on the cloud amount image information in the sampling database to obtain geometric quality inspection results and geometric quality inspection image information.

[0062] In the above steps, geometric accuracy inspection can be to generate checkpoints by comparing the cloud amount image information in the sampling database with reference data, and perform matching calculations to calculate the error results, so as to obtain the corresponding geometric quality inspection results and geometric quality inspection image information.

[0063] Exemplarily, geometric quality inspection can include quality inspections such as geometric positioning accuracy, fusion registration accuracy, spatial reference, and edge-matching accuracy. Geometric quality inspection can also include data geometric plane accuracy inspection, which includes: (1) Spatial reference: the correctness of information such as the coordinate system, elevation datum, and map projection parameters of the image; (2) Geometric plane positioning accuracy, panchromatic-multispectral registration accuracy, and multispectral spectral inter-registration accuracy; (3) Satellite attitude parameters (roll angle and yaw angle).

[0064] Please refer to Figure 2 , the steps of performing geometric quality detection on the cloud amount image information to obtain geometric quality detection image information specifically include:

[0065] Step S5-1: Perform similarity encoding on the cloud amount image information in the sampling database and the preset reference image information respectively to obtain cloud amount image encoding and reference image encoding.

[0066] In the above steps, by respectively performing similarity encoding on the cloud cover image information in the sampling database and the preset reference image information, the obtained cloud cover image encoding and reference image encoding can be used to provide data support for subsequent similarity judgment, so as to be used to judge the geometric quality of the cloud cover image information.

[0067] Exemplarily, it can be encoded by hash encoding, so that it can be dimensionally reduced from high-dimensional data to low-dimensional data without changing the similarity between the cloud cover image information in the original sampling database and the preset reference image information, thereby effectively reducing the memory required for its operation and improving the processing efficiency while ensuring the accuracy of similarity calculation.

[0068] Step S5-2: Calculate the similarity distance between the cloud cover image encoding and the reference image encoding by using a distance calculation method.

[0069] In the above steps, the similarity distance calculation method can be selected diversely according to the actual situation, such as similarity distance calculation methods in existing technologies such as Euclidean similarity distance, Mahalanobis similarity distance, and Chebyshev similarity distance. Without limiting the specific calculation method on the premise of being able to implement the present invention.

[0070] Step S5-3: Compare the similarity distance with the preset similarity threshold, and generate a geometric quality detection image according to the obtained judgment result.

[0071] In the above steps, by comparing the similarity distance with the preset similarity threshold, the magnitude of the similarity can be judged, so as to obtain the corresponding geometric quality detection image. The specific value of the similarity threshold can be set according to the actual accuracy requirement.

[0072] Step S6: Perform radiation quality inspection on the geometric quality inspection image information to obtain a radiation quality inspection result.

[0073] Please refer to Figure 1 , the above radiation quality inspection includes at least one of inter-slice radiation anomaly detection, inter-slice color difference anomaly detection, color cast detection, missing detection, garbled detection, or stripe noise detection.

[0074] In the above steps, by performing radiation quality inspections such as inter-slice radiation anomaly detection, inter-slice color difference anomaly detection, color cast detection, missing detection, garbled detection, and stripe noise detection on the geometric quality detection image information, the final product quality detection result can be accurately obtained.

[0075] Please refer to Figure 3 , the above inter-slice radiation anomaly detection specifically includes:

[0076] Step S6-1: Calculate the gradient values of the projections of all image layers in the geometric quality inspection image information respectively, and obtain the image layer gradient value information corresponding to each layer.

[0077] In the above steps, there will be color abnormalities between or at the splicing points of the images between slices in the geometric quality inspection image information, and the number of CCD slices spliced by different satellite images is also inconsistent. There must be splicing points of images at all intervals between CCD slices, that is, the longitudinal stripes therein belong to edge mutations. By analyzing the gradient changes of the longitudinal projection, the stripes and their positions can be perceived. Therefore, first, by accumulating the longitudinal projections of each channel of the geometric quality inspection image information, the gradient values of the projections of each channel are calculated, which can provide data support for subsequent analysis of the gradient changes of the longitudinal projection.

[0078] Step S6-2: Obtain the first gradient maximum value and the first gradient mean value in the image layer gradient value information corresponding to each layer respectively, and record the ratio of the first gradient maximum value to the first gradient mean value as r1.

[0079] Step S6-3: Perform edge detection on the projections of all image layers in the geometric quality inspection image information respectively, and obtain the edge detection images corresponding to each layer.

[0080] In the above steps, through edge detection, the required boundary and shape information in the projections of all image layers in the geometric quality inspection image information can be identified, so as to perform analysis and calculation on it.

[0081] Exemplarily, Canny Edge Detection algorithm can be used for edge detection. Among them, the Canny Edge Detection algorithm specifically includes the following steps: 1. Gaussian filtering; 2. Calculate the gradient value and gradient direction; 3. Filter non-maximum values; 4. Use upper and lower thresholds to detect edges. By processing all image layers through Gaussian filtering, a denoised image layer is obtained, which can make the original image smoother and may increase the width of the edge at the same time. Then, through the calculated gradient value and gradient direction of the denoised image layer, the change degree and direction of the gray value of the denoised image layer can be identified. Since the edge may be magnified during the Gaussian filtering process. Then, through the step of filtering non-maximum values, a rule is used to filter out the points that are not edges, so that the width of the edge is as small as possible, preferably 1 pixel. Finally, by using upper and lower thresholds to detect edges, the required boundary and shape information can be clearly defined.

[0082] Step S6-4: Obtain the second gradient maximum value and the second gradient mean value in the edge detection image corresponding to each layer respectively, and record the ratio of the second gradient maximum value to the second gradient mean value as r2.

[0083] Step S6-5: Determine the size relationship between r1 and r2 of the corresponding layer. If r1 ≥ r2, it is recorded that there is no inter-chip radiation anomaly problem. If r1 < r2, it is recorded that there is an inter-chip radiation anomaly problem.

[0084] In the above steps, by judging the size relationship between r1 and r2 of the corresponding layer, it is possible to clearly understand whether there is an inter-chip radiation anomaly problem.

[0085] Please refer to Figure 1 , the above color cast detection is the automatic detection of color cast images under CIE Lab.

[0086] In the above steps, color cast refers to the abnormal phenomenon that the radiation energy responses of one or several bands of an image are inconsistent, presenting a color tone deviation of the overall image towards the band with high radiation energy response. Color cast may occur in the process of satellite image shooting due to light or angle problems. In the existing technology, the RGB color space is the simplest color space, but the biggest limitation of the RGB color space is that when using the Euclidean distance to describe the difference between two colors, the calculated distance between the two colors cannot correctly represent the true difference between the two colors actually perceived by people. By using the CIE Lab color space, the distance between the colors calculated in this space is basically consistent with the actual perception difference. Its histogram can more objectively reflect the degree of image color cast, so the automatic detection of color cast images under CIELab will be more accurate and effective.

[0087] Please refer to Figure 4 , the above missing detection specifically includes:

[0088] Step S6-6: Obtain the double-peak histogram image information of all frames in the geometric quality inspection image information.

[0089] In the above steps, missing means that there is no texture information or lack of a certain band in some rows and columns of the image; the radiation values of one or several bands in all or part of the image area are abnormal; during the transmission or shooting of remote sensing images, information loss occurs in some bands, resulting in "black stripes" in the image. The pixel values of the missing part are not necessarily 0, but some points with relatively small pixel values. The adaptive threshold is used to segment the image of the double-peak histogram, and the area below the threshold is considered the missing part. Through the fast non-local means denoising method, while taking into account noise suppression and texture retention, edge detection is performed on it to detect the boundaries of the block areas, so that missing detection can be carried out. Therefore, first obtain the double-peak histogram image information of all frames in the geometric quality detection image information, which can facilitate subsequent image segmentation processing.

[0090] Step S6-7: Segment the dual-peak histogram image information using the adaptive threshold method. If it is lower than the adaptive threshold, the block area is marked as a missing area.

[0091] In the above steps, by segmenting the dual-peak histogram image information using the adaptive threshold method, the missing areas can be well marked out.

[0092] Exemplarily, if the gray-level histogram is significantly bimodal, the gray level corresponding to the valley between the two peaks can be selected as the adaptive threshold.

[0093] Step S6-8: Use the fast non-local means denoising method to perform edge detection on the missing area to obtain the boundary information of the missing area.

[0094] In the above steps, by using the fast non-local means denoising method to perform edge detection on the missing area, that is, by using the method of smoothing and taking the mean in the area around a target pixel to perform edge detection on it. Since non-local means filtering means that it uses all the pixels in the missing area, and these pixels are weighted and averaged according to a certain similarity, the clarity of the filtered image is high and no details are lost.

[0095] Step S6-9: Generate missing detection result information using the boundary information of the missing area.

[0096] Step S7: Obtain the final inspection result using the specification inspection result, cloud amount calculation result, geometric quality inspection result, and radiation quality inspection result.

[0097] In the above steps, by comprehensively considering the specification inspection result, cloud amount calculation result, geometric quality inspection result, and radiation quality inspection result, a more rich and diverse final inspection result can be obtained, and various data can be used for mutual verification, thereby further improving the accuracy of the inspection result.

[0098] It should be noted that the multi-source satellite remote sensing image quality inspection method proposed in the present invention can be implemented through an application program, so as to realize the automatic quality inspection of multi-source satellite remote sensing images. It can weaken the influence of manual operations on the entire inspection process, so that the quality of a large number of multi-source satellite remote sensing images can be inspected with higher efficiency and quality, which has great practical and social significance for the proposal and improvement of the quality inspection method of a large number of multi-source satellite remote sensing images.

[0099] Please refer to Figure 5 , Figure 5A block diagram of an electronic device provided in an embodiment of the present invention. The electronic device includes a memory 101, a processor 102 and a communication interface 103, and the memory 101, the processor 102 and the communication interface 103 are electrically connected to each other directly or indirectly to realize data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The memory 101 can be used to store software programs and modules, such as program instructions / modules corresponding to a method for quality inspection of a massive multi-source satellite remote sensing image product provided in an embodiment of the present application, and the processor 102 executes various functional applications and data processing by executing the software programs and modules stored in the memory 101. The communication interface 103 can be used to communicate signaling or data with other node devices.

[0100] Among them, the memory 101 can be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc.

[0101] The processor 102 may be an integrated circuit chip with signal processing capability. The processor 102 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0102] Understandably, Figure 5 The structure shown is for illustration only. The electronic device may also include Figure 5 More or fewer components as shown, or with Figure 5 Different configurations shown. Figure 5 Each component shown in the figure can be implemented by hardware, software or a combination thereof.

[0103] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions, and operations of devices, methods, and computer program products according to multiple embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0104] In addition, in each embodiment of the present application, the functional modules can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part.

[0105] If the above functions are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, 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 each embodiment of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0106] The above are only the preferred embodiments of the present application and are not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0107] For those skilled in the art, it is obvious that the present application is not limited to the details of the above-mentioned exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or basic characteristics of the present application. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present application is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present application. Any reference signs in the claims should not be construed as limiting the claims involved.

Claims

1. An automated quality inspection method for a large amount of multi-source satellite remote sensing images, characterized in that, it includes the following steps: Obtain multi-source satellite remote sensing images to obtain original image information; Conduct a specification inspection on the original image information to obtain a specification inspection result; Conduct cloud amount calculation on the original image information after the specification inspection to obtain a cloud amount calculation result and cloud amount image information including cloud amount proportion information; Sample the cloud amount image information according to preset conditions to obtain a sampling database; Conduct geometric quality inspection on the cloud amount image information in the sampling database to obtain a geometric quality inspection result and geometric quality inspection image information; Conduct radiometric quality inspection on the geometric quality inspection image information to obtain a radiometric quality inspection result. The radiometric quality inspection includes inter-slice radiometric anomaly detection, inter-slice color difference anomaly detection, color cast detection, missing detection, garbled code detection, and stripe noise detection. The specific inter-slice radiometric anomaly detection includes: respectively calculating the gradient values of the projections of all image layers in the geometric quality inspection image information to obtain the image layer gradient value information of the corresponding layers; respectively obtaining the first gradient maximum value and the first gradient average value in the image layer gradient value information of the corresponding layers, and recording the ratio of the first gradient maximum value to the first gradient average value as r1; respectively performing edge detection on the projections of all image layers in the geometric quality inspection image information to obtain the edge detection images of the corresponding layers; respectively obtaining the second gradient maximum value and the second gradient average value in the edge detection images of the corresponding layers, and recording the ratio of the second gradient maximum value to the second gradient average value as r2; respectively judging the size relationship between r1 and r2 of the corresponding layers. If r1≥r2, it is recorded that there is no inter-slice radiometric anomaly problem. If r1<r2, it is recorded that there is an inter-slice radiometric anomaly problem; Use the specification inspection result, cloud amount calculation result, geometric quality inspection result, and radiometric quality inspection result to obtain a final inspection result.

2. An automated quality inspection method for a large amount of multi-source satellite remote sensing images according to claim 1, characterized in that, The specific specification inspection of the original image information includes at least one of metadata quality detection, file integrity detection, data consistency detection, data standardization detection, or data validity detection.

3. An automated quality inspection method for a large amount of multi-source satellite remote sensing images according to claim 1, characterized in that, The step of conducting geometric quality inspection on the cloud amount image information in the sampling database to obtain a geometric quality inspection result and geometric quality inspection image information specifically includes: Perform similarity coding on the cloud amount image information in the sampling database and the preset reference image information respectively to obtain a cloud amount image code and a reference image code; Use a distance calculation method to calculate the similarity distance between the cloud amount image code and the reference image code; Compare the similarity distance with the preset similarity threshold, and generate a geometric quality detection image according to the obtained judgment result.

4. An automated quality inspection method for a large amount of multi-source satellite remote sensing images according to claim 1, characterized in that, The color cast detection is to automatically detect color cast images under CIE Lab.

5. A method for automated quality inspection of massive multi-source satellite remote sensing images according to claim 1, characterized in that, the missing detection specifically includes: obtaining the double-peak histogram image information of all frames in the geometric quality inspection image information; segmenting the double-peak histogram image information by using an adaptive threshold method, and if it is lower than the adaptive threshold, the block area is marked as a missing area; performing edge detection on the missing area by using a fast non-local mean denoising method to obtain the boundary information of the missing area; generating missing detection result information by using the boundary information of the missing area.

6. An electronic device, characterized in that, it includes: a memory for storing one or more programs; a processor; when the one or more programs are executed by the processor, the method according to any one of claims 1-5 is implemented.

7. A computer-readable storage medium, on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the method according to any one of claims 1-5 is implemented.

Images