Multi-load large-width image data satellite-borne real-time registration method and system

Abstract

The invention provides a multi-load large-width image data satellite-borne real-time registration method and system, and the method comprises the steps: receiving image data transmitted by multiple loads, and carrying out the data analysis work; enabling the ZYNQ to receive the multi-load image data and the auxiliary data, and respectively carrying out caching processing; carrying out four-corner-point pixel coarse positioning operation on cached head and tail data of the load; calculating the maximum and minimum values of the longitude and latitude; extracting corresponding DEM (Digital Elevation Model) equal-angle grid data; frame-by-frame pixel-by-pixel control-point-free coarse positioning processing is carried out, elevation data of a pixel point is taken out from DDR, and elevation-constrained fine positioning iterative solution is carried out; dEM grid projection processing is carried out on multi-load fine positioning calculation results respectively; carrying out traversal interpolation processing; subsequent multi-load multi-channel combined treatment is carried out; and repeating until the transmission of the load data is stopped or the calculation of the cache data is completed. The invention provides in-orbit multi-load data real-time gridding registration processing, and solves the current situation that multi-load observation is carried out in different domains at the same time.

Description

Technical Field

[0001] The present invention relates to the field of embedded image real-time preprocessing, and in particular, to a satellite-based real-time registration method and system for multi-payload and wide-width image data; more particularly, to a satellite-based real-time registration method and system for multi-payload and wide-width image data based on global DEM data. Background Art

[0002] Remote sensing is an important application of space information network, playing an important role in many fields such as environment, transportation, ocean, agriculture, water conservancy, surveying and mapping, geology, etc. With the development of business needs and the increase in demand for different detection spectrum data, satellites are developing in the direction of multi-payload and multi-spectral band. For example, my country's FY-3 series of satellites have more than ten kinds of payloads and thousands of spectrum bands. However, with the continuous increase in the number of payloads and spectrum bands, the amount of data generated is also growing exponentially. While the massive multi-spectral satellite remote sensing data brings a higher quality data source to remote sensing applications, it also reduces the timeliness of data application.

[0003] In actual application, considering the real-time and application timeliness issues, the ground wants to quickly obtain remote sensing products of the observation area, rather than waiting for the original data on the satellite to be transmitted and then conduct ground post-processing. Therefore, onboard real-time payload data processing has become a pain point that needs to be solved urgently. At present, wide-bandwidth satellites mostly use linear array CCD splicing or multi-detector scanning to achieve wide-bandwidth requirements. Considering that the imaging mechanisms of different payload designs are different, the multi-spectral data are not based on the same grid reference during actual imaging detection. This multi-payload imaging mechanism brings certain difficulties to the onboard real-time multi-spectral data fusion processing.

[0004] First, there are many payloads and spectral bands. When different spectral bands are observed at the same time, the observation points are not on the same grid reference, which makes it difficult to jointly use multiple spectral bands. Second, the onboard real-time positioning processing adopts a point-free positioning method, which lacks elevation constraints and cannot effectively guarantee the positioning accuracy. Third, the massive amount of observation data cannot be cached first and then slowly post-processed like ground processing. Due to the limitation of onboard cache resources, real-time positioning solution and alignment are required, which poses a challenge to the onboard processing strategy.

[0005] In view of the difficulty of onboard real-time image data registration caused by the characteristics of multi-payload wide-width imaging, a method for onboard real-time registration of multi-payload wide-width image data based on global DEM data is proposed. This method mainly stores global DEM grid data in the form of an onboard solid-state disk plug-in, and then quickly solves the latitude and longitude coordinate range of the observation data based on the observation data within a certain period of time. According to the appropriately expanded range of longitude and latitude coordinates, the DEM grid data within the observation range is obtained from the solid-state disk and cached. Finally, the observation data is precisely positioned pixel by pixel with elevation constraints, and the cached DEM grid data is processed with equal-angle projection and traversal interpolation to obtain the registered grid data of multiple payloads and multiple spectral bands for the back-end to carry out multi-spectral joint processing. Solve the problem of onboard real-time registration of wide-width multi-payload image data. Summary of the invention

[0006] In view of the defects in the prior art, the object of the present invention is to provide a

[0007] A method for real-time satellite registration of multi-payload wide-width image data provided by the present invention comprises:

[0008] Step S1: The interface module FPGA receives the image data transmitted by multiple payloads, performs data analysis, extracts the image data and auxiliary data of multiple payloads, repackages them and sends them to ZYNQ;

[0009] Step S2: Instruct ZYNQ to receive multi-payload image data and auxiliary data, and perform cache processing respectively;

[0010] Step S3: When the image data cache meets the preset amount, a four-corner pixel coarse positioning operation is performed on the cached head and tail data of any load;

[0011] Step S4: Calculate the maximum and minimum values ​​of longitude and latitude for the rough positioning results of the four corner point pixels, obtain the geographic coordinate range of the observation data and expand it;

[0012] Step S5: extracting corresponding DEM equal-angle grid data in the external storage module according to the geographic coordinates of the observed data, and storing them in the DDR cache;

[0013] Step S6: Performing coarse positioning processing on the observation data frame by frame and pixel by pixel without control points, and taking out the elevation data of the pixel point in DDR according to the latitude and longitude results of the coarse positioning, performing iterative solution of precise positioning with elevation constraints, and independently caching the precise positioning solution results of multiple loads;

[0014] Step S7: Perform DEM grid projection processing on the precise positioning solution results of multiple loads respectively;

[0015] Step S8: Perform traversal interpolation processing on the data after DEM grid projection of multiple loads to form complete DEM grid data;

[0016] Step S9: sending the interpolated data to the processing backend for subsequent multi-load and multi-channel joint processing;

[0017] Repeat steps S1 to S9 until the payload data stops transmitting or the cache data is resolved.

[0018] Preferably, in step S2:

[0019] The cache process uses a circular queue to store data, taking into account both the stability of memory usage in the onboard software and the convenience of development and use.

[0020] Preferably, in step S4:

[0021] The geographic coordinate range calculation uses the four-corner point coarse positioning method to frame the latitude and longitude grid range of the observation data, reducing the computing resource overhead caused by pixel-by-pixel positioning operations.

[0022] Preferably, in step S5:

[0023] DEM data extraction adopts the design of satellite-borne external storage module, and the communication method adopts optical fiber interface to meet the needs of DEM data storage and regional DEM data extraction;

[0024] DEM data storage is achieved by using a solid-state disk plug-in. The DEM grid data resolution is 0.0005°×0.0005°. It is designed to be programmable on-orbit and can replace DEM data in any area to meet the needs of on-orbit dynamic updates.

[0025] Preferably, in step S6:

[0026] Based on the pixel precise positioning calculation constrained by DEM elevation information, the influence of elevation on positioning accuracy is comprehensively considered to make up for the defect of insufficient plane positioning accuracy without control points;

[0027] The pixel precise positioning operation constrained by DEM elevation information adopts the method of pre-coarse positioning to extract DEM grid data within the observation area and store it in the DDR of the processing board, avoiding the resource overhead caused by repeated reading of the solid-state disk and improving the speed of DEM elevation data acquisition.

[0028] Preferably, in step S7:

[0029] The precise positioning solution results of multiple loads are projected onto DEM grids, and the grid data after rough projection is subjected to another traversal interpolation process to obtain complete multi-load grid data based on DEM registration;

[0030] Real-time grid registration processing is performed on different types of payload data carried by satellites in orbit to solve the problem of multiple payloads observing different domains at the same time. The processed multi-payload grid data can be sent to the back end for different types of band combination interpretation processing.

[0031] Preferably, a satellite-borne embedded architecture of a combination of FPGA, ZYNQ and solid-state disk storage is provided, which adopts a preset high-throughput FPGA and a preset high-performance ARM processor, as well as a preset large-capacity and high-throughput solid-state disk combination to meet the requirements of large real-time image throughput and high computing resources. At the same time, the solid-state disk storage method is also used to obtain the DEM grid data of the observation area in real time and carry out pixel-by-pixel geometric precision positioning processing of multiple payloads.

[0032] A multi-payload wide-width image data satellite-borne real-time registration system provided by the present invention comprises:

[0033] Module M1: The interface module FPGA receives the image data transmitted by multiple payloads, performs data analysis, extracts the image data and auxiliary data of multiple payloads, repackages them and sends them to ZYNQ;

[0034] Module M2: Instructs ZYNQ to receive multi-payload image data and auxiliary data and perform cache processing respectively;

[0035] Module M3: When the image data cache meets the preset amount, a four-corner pixel coarse positioning operation is performed on the cache head and tail data of any load;

[0036] Module M4: Calculate the maximum and minimum values ​​of longitude and latitude for the rough positioning results of the four corner point pixels, obtain the geographic coordinate range of the observation data and expand it;

[0037] Module M5: extracts the corresponding DEM equiangular grid data in the external storage module according to the geographic coordinates of the observed data, and stores it in the DDR cache;

[0038] Module M6: Carry out coarse positioning processing of observation data frame by frame and pixel by pixel without control points. At the same time, according to the latitude and longitude results of coarse positioning, the elevation data of the pixel point is taken out from DDR, and the precise positioning iterative solution of elevation constraints is carried out. The precise positioning solution results of multiple loads are cached independently.

[0039] Module M7: DEM grid projection processing is performed on the precise positioning solution results of multiple loads;

[0040] Module M8: Perform traversal interpolation processing on the DEM grid projection data of multiple loads to form complete DEM grid data;

[0041] Module M9: Packaging the interpolated data to the processing backend for subsequent multi-load and multi-channel joint processing;

[0042] Modules M1 to M9 are triggered repeatedly until the payload data stops transmitting or the buffer data is resolved.

[0043] Preferably, in the module M2:

[0044] The cache process uses a circular queue to store data, taking into account the stability of memory usage in the onboard software and the convenience of development and use;

[0045] In the module M4:

[0046] The geographic coordinate range calculation uses the four-corner point coarse positioning method to frame the latitude and longitude grid range of the observation data, reducing the computing resource overhead caused by pixel-by-pixel positioning operations;

[0047] In the module M5:

[0048] DEM data extraction adopts the design of satellite-borne external storage module, and the communication method adopts optical fiber interface to meet the needs of DEM data storage and regional DEM data extraction;

[0049] DEM data storage is implemented by using a solid-state disk plug-in. The DEM grid data resolution is 0.0005°×0.0005°. The design is programmable on-orbit, and the DEM data of any area can be replaced to meet the needs of on-orbit dynamic updates.

[0050] In the module M6:

[0051] Based on the pixel precise positioning calculation constrained by DEM elevation information, the influence of elevation on positioning accuracy is comprehensively considered to make up for the defect of insufficient plane positioning accuracy without control points;

[0052] The pixel precise positioning operation constrained by DEM elevation information adopts the method of pre-coarse positioning to extract DEM grid data within the observation area and store it in the DDR of the processing board, avoiding the resource overhead caused by repeated reading of the solid-state disk and improving the speed of DEM elevation data acquisition.

[0053] Preferably, in the module M7:

[0054] The precise positioning solution results of multiple loads are projected onto DEM grids, and the grid data after rough projection is subjected to another traversal interpolation process to obtain complete multi-load grid data based on DEM registration;

[0055] Carry out real-time grid registration processing on different types of payload data carried by satellites in real time on orbit to solve the current situation of multiple payloads observing different domains at the same time. The processed multi-payload grid data can be sent to the back-end for different types of band combination interpretation processing;

[0056] The invention provides a satellite-borne embedded architecture of a combination of FPGA, ZYNQ and SSD storage, which adopts a preset high-throughput FPGA and a preset high-performance ARM processor, as well as a preset large-capacity and high-throughput SSD combination to meet the requirements of large real-time image throughput and high computing resources. At the same time, the SSD storage method is also used to obtain the DEM grid data of the observation area in real time and carry out pixel-by-pixel geometric precise positioning processing of multiple payloads.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. The present invention provides a satellite-borne embedded architecture of FPGA+ZYNQ+SSD storage combination, which adopts high-throughput FPGA and high-performance ARM processor, as well as large-capacity and high-throughput SSD combination, which not only meets the requirements of high real-time image throughput and high computing resources, but also meets the requirements of large-capacity DEM data storage;

[0059] 2. The present invention proposes a method for quickly obtaining the latitude and longitude range of the observation area by coarse positioning of corner points, which effectively reduces the computing resource overhead caused by the pixel-by-pixel coarse positioning operation when obtaining the observation range;

[0060] 3. The present invention proposes a real-time gridding registration processing of on-orbit multi-payload data to solve the current situation of simultaneous observation of multiple payloads in different domains. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0062] Figure 1 This is a flow chart of the real-time satellite registration method for multi-payload wide-width image data based on global DEM data;

[0063] Figure 2 This is an example diagram of a circular queue-based data cache. DETAILED DESCRIPTION

[0064] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0065] Embodiment 1:

[0066] The present invention relates to a satellite-borne real-time registration method for multi-payload wide-width image data based on global DEM data, which provides a solution to the real-time registration of multi-payload data, reduces the consumption of embedded resources and solves the problem that multi-channel image data cannot be jointly applied in real time due to simultaneous observation of multiple payloads in different domains; at the same time, a satellite-borne embedded architecture of FPGA+ZYNQ+solid-state disk storage combination is provided, which not only meets the problem of real-time processing of satellite-borne large data volumes, but also meets the demand for satellite-borne large-capacity data storage.

[0067] According to the present invention, a multi-payload wide-width image data on-board real-time registration method is provided. Figure 1-Figure 2 As shown, including:

[0068] Step S1: The interface module FPGA receives the image data transmitted by multiple payloads, performs data analysis, extracts the image data and auxiliary data of multiple payloads, repackages them and sends them to ZYNQ;

[0069] Step S2: Instruct ZYNQ to receive multi-payload image data and auxiliary data, and perform cache processing respectively;

[0070] Specifically, in step S2:

[0071] The cache process uses a circular queue to store data, taking into account both the stability of memory usage in the onboard software and the convenience of development and use.

[0072] Step S3: When the image data cache meets the preset amount, a four-corner pixel coarse positioning operation is performed on the cached head and tail data of any load;

[0073] Step S4: Calculate the maximum and minimum values ​​of longitude and latitude for the rough positioning results of the four corner point pixels, obtain the geographic coordinate range of the observation data and expand it;

[0074] Specifically, in step S4:

[0075] The geographic coordinate range calculation uses the four-corner point coarse positioning method to frame the latitude and longitude grid range of the observation data, reducing the computing resource overhead caused by pixel-by-pixel positioning operations.

[0076] Step S5: extracting corresponding DEM equal-angle grid data in the external storage module according to the geographic coordinates of the observed data, and storing them in the DDR cache;

[0077] Specifically, in step S5:

[0078] DEM data extraction adopts the design of satellite-borne external storage module, and the communication method adopts optical fiber interface to meet the needs of DEM data storage and regional DEM data extraction;

[0079] DEM data storage is achieved by using a solid-state disk plug-in. The DEM grid data resolution is 0.0005°×0.0005°. It is designed to be programmable on-orbit and can replace DEM data in any area to meet the needs of on-orbit dynamic updates.

[0080] Step S6: Performing coarse positioning processing on the observation data frame by frame and pixel by pixel without control points, and taking out the elevation data of the pixel point in DDR according to the latitude and longitude results of the coarse positioning, performing iterative solution of precise positioning with elevation constraints, and independently caching the precise positioning solution results of multiple loads;

[0081] Specifically, in step S6:

[0082] Based on the pixel precise positioning calculation constrained by DEM elevation information, the influence of elevation on positioning accuracy is comprehensively considered to make up for the defect of insufficient plane positioning accuracy without control points;

[0083] The pixel precise positioning operation constrained by DEM elevation information adopts the method of pre-coarse positioning to extract DEM grid data within the observation area and store it in the DDR of the processing board, avoiding the resource overhead caused by repeated reading of the solid-state disk and improving the speed of DEM elevation data acquisition.

[0084] Step S7: Perform DEM grid projection processing on the precise positioning solution results of multiple loads respectively;

[0085] Specifically, in step S7:

[0086] The precise positioning solution results of multiple loads are projected onto DEM grids, and the grid data after rough projection is subjected to another traversal interpolation process to obtain complete multi-load grid data based on DEM registration;

[0087] Real-time grid registration processing is performed on different types of payload data carried by satellites in orbit to solve the problem of multiple payloads observing different domains at the same time. The processed multi-payload grid data can be sent to the back end for different types of band combination interpretation processing.

[0088] Step S8: Perform traversal interpolation processing on the data after DEM grid projection of multiple loads to form complete DEM grid data;

[0089] Step S9: sending the interpolated data to the processing backend for subsequent multi-load and multi-channel joint processing;

[0090] Repeat steps S1 to S9 until the payload data stops transmitting or the cache data is resolved.

[0091] Specifically, a satellite-borne embedded architecture of a combination of FPGA, ZYNQ and solid-state disk storage is provided, which adopts a preset high-throughput FPGA and a preset high-performance ARM processor, as well as a preset large-capacity and high-throughput solid-state disk combination to meet the requirements of large real-time image throughput and high computing resources. At the same time, the solid-state disk storage method is also used to obtain the DEM grid data of the observation area in real time, and carry out pixel-by-pixel geometric precise positioning processing of multiple payloads.

[0092] Embodiment 2:

[0093] Embodiment 2 is a preferred example of Embodiment 1, and is used to illustrate the present invention in more detail.

[0094] The present invention also provides a satellite-borne real-time registration system for multi-payload and wide-width image data. The satellite-borne real-time registration system for multi-payload and wide-width image data can be implemented by executing the process steps of the satellite-borne real-time registration method for multi-payload and wide-width image data, that is, those skilled in the art can understand the satellite-borne real-time registration method for multi-payload and wide-width image data as a preferred implementation of the satellite-borne real-time registration system for multi-payload and wide-width image data.

[0095] A multi-payload wide-width image data satellite-borne real-time registration system provided by the present invention comprises:

[0096] Module M1: The interface module FPGA receives the image data transmitted by multiple payloads, performs data analysis, extracts the image data and auxiliary data of multiple payloads, repackages them and sends them to ZYNQ;

[0097] Module M2: Instructs ZYNQ to receive multi-payload image data and auxiliary data and perform cache processing respectively;

[0098] Specifically, in the module M2:

[0099] The cache process uses a circular queue to store data, taking into account the stability of memory usage in the onboard software and the convenience of development and use;

[0100] Module M3: When the image data cache meets the preset amount, a four-corner pixel coarse positioning operation is performed on the cache head and tail data of any load;

[0101] Module M4: Calculate the maximum and minimum values ​​of longitude and latitude for the rough positioning results of the four corner point pixels, obtain the geographic coordinate range of the observation data and expand it;

[0102] In the module M4:

[0103] The geographic coordinate range calculation uses the four-corner point coarse positioning method to frame the latitude and longitude grid range of the observation data, reducing the computing resource overhead caused by pixel-by-pixel positioning operations;

[0104] Module M5: extracts the corresponding DEM equiangular grid data in the external storage module according to the geographic coordinates of the observed data, and stores it in the DDR cache;

[0105] In the module M5:

[0106] DEM data extraction adopts the design of satellite-borne external storage module, and the communication method adopts optical fiber interface to meet the needs of DEM data storage and regional DEM data extraction;

[0107] DEM data storage is implemented by using a solid-state disk plug-in. The DEM grid data resolution is 0.0005°×0.0005°. The design is programmable on-orbit, and the DEM data of any area can be replaced to meet the needs of on-orbit dynamic updates.

[0108] Module M6: Carry out coarse positioning processing of observation data frame by frame and pixel by pixel without control points. At the same time, according to the latitude and longitude results of coarse positioning, the elevation data of the pixel point is taken out from DDR, and the precise positioning iterative solution of elevation constraints is carried out. The precise positioning solution results of multiple loads are cached independently.

[0109] In the module M6:

[0110] Based on the pixel precise positioning calculation constrained by DEM elevation information, the influence of elevation on positioning accuracy is comprehensively considered to make up for the defect of insufficient plane positioning accuracy without control points;

[0111] The pixel precise positioning operation constrained by DEM elevation information adopts the method of pre-coarse positioning to extract DEM grid data within the observation area and store it in the DDR of the processing board, avoiding the resource overhead caused by repeated reading of the solid-state disk and improving the speed of DEM elevation data acquisition.

[0112] Module M7: DEM grid projection processing is performed on the precise positioning solution results of multiple loads;

[0113] Specifically, in the module M7:

[0114] The precise positioning solution results of multiple loads are projected onto DEM grids, and the grid data after rough projection is subjected to another traversal interpolation process to obtain complete multi-load grid data based on DEM registration;

[0115] Carry out real-time grid registration processing on different types of payload data carried by satellites in real time on orbit to solve the current situation of multiple payloads observing different domains at the same time. The processed multi-payload grid data can be sent to the back-end for different types of band combination interpretation processing;

[0116] Module M8: Perform traversal interpolation processing on the DEM grid projection data of multiple loads to form complete DEM grid data;

[0117] Module M9: Packaging the interpolated data to the processing backend for subsequent multi-load and multi-channel joint processing;

[0118] Modules M1 to M9 are triggered repeatedly until the payload data stops transmitting or the buffer data is resolved.

[0119] The invention provides a satellite-borne embedded architecture of a combination of FPGA, ZYNQ and SSD storage, which adopts a preset high-throughput FPGA and a preset high-performance ARM processor, as well as a preset large-capacity and high-throughput SSD combination to meet the requirements of large real-time image throughput and high computing resources. At the same time, the SSD storage method is also used to obtain the DEM grid data of the observation area in real time and carry out pixel-by-pixel geometric precise positioning processing of multiple payloads.

[0120] Embodiment 3:

[0121] Embodiment 3 is a preferred example of Embodiment 1, and is used to illustrate the present invention in more detail.

[0122] The purpose of the present invention is to address the deficiencies in the prior art and provide a practical satellite-borne real-time registration method for multi-payload wide-width image data based on global DEM data. The method uses coarse corner point positioning to quickly obtain the latitude and longitude coordinate range of the observation data, and then performs real-time DEM grid projection processing on the pixel-by-pixel positioning results under elevation constraints. This can reduce the massive data caching and processing pressure brought by multi-payload and multi-spectral bands to the greatest extent, and meet the needs of satellite-borne real-time and fast multi-payload image registration.

[0123] This embodiment provides a method for real-time satellite registration of multi-payload wide-width image data based on global DEM data, comprising the following steps:

[0124] (1) The interface module FPGA receives the wide-width image data transmitted by multiple payloads, performs data analysis, extracts the image data and auxiliary data of the multiple payloads, repackages them, and sends them to ZYNQ (an extensible processing platform of Xilinx, one of whose chip series is called ZYNQ, which is a general abbreviation in the embedded processing industry); in order to better illustrate the implementation and operation process of the present invention, the number of payloads is simplified to N here;

[0125] (2) ZYNQ receives multi-payload image data and auxiliary data and performs cache processing respectively; ZYNQ caches various types of data in a circular queue, which can meet the stability of memory usage in the onboard software and the convenience of development and use. Figure 2 It can better help explain the storage method of this step;

[0126] (3) ZYNQ When the image data cache meets a certain value, a four-corner pixel coarse positioning operation is performed on the cache head and tail data of any load therein; in order to better illustrate the implementation mode and operation process of the present invention, the certain value in which the image data cache meets a certain value is set as Period, and any load in the cache head and tail data of any load is assumed to be load 1, and the four corner points are the first and last observation points of the first frame and the Period frame of the image cache of load 1, which are counted as pt1~pt4; a pixel coarse positioning operation is performed on pt1~pt4 to obtain (lon1, lat1), (lon2, lat2), (lon3, lat3) and (lon4, lat4).

[0127] (4) ZYNQ calculates the maximum and minimum values ​​of longitude and latitude for the rough positioning results of pt1 to pt4, that is, it calculates the maximum and minimum values ​​of lon and the maximum and minimum values ​​of lat of (lon1, lat1), (lon2, lat2), (lon3, lat3) and (lon4, lat4), which are recorded as lon min ,lon max ,lat min ,lat max , and expand Δt (angle, configurable parameters) to obtain lon1, lon2, lat1 and lat2;

[0128] (5) ZYNQ extracts the corresponding DEM equiangular grid data from the external storage module according to the geographic coordinates of the observation data, and stores it in the DDR cache; in order to better illustrate the implementation and operation process of the present invention, the resolution of the global DEM grid data is set to 0.0005° (the parameter is configurable), that is, the size of the global DEM grid data is 720000*360000; the local DEM grid data of the observation area is framed in the global DEM grid data according to lon1, lon2, lat1 and lat2, and the size of the local DEM grid data is calculated as: (lon2-lon1) / 0.005, (lat2-lat1) / 0.005, which are recorded as Width and Height respectively;

[0129] (6) ZYNQ performs frame-by-frame and pixel-by-pixel coarse positioning processing on the observation data without control points. In order to better illustrate the implementation and operation process of the present invention, one of the observation pixels ptt of load 1 is taken as an example to calculate the coarse positioning longitude and latitude results (lont, latt) of ptt. The elevation data of the observation pixel ptt at the position (indi, indj) in the DEM grid of the observation area is obtained according to indi=(lont-lon1) / 0.005, indj=(latt-lat1) / 0.005, and the elevation constraint is implemented. Iterative solution of precise positioning is performed. The precise positioning solution of all pixels in the observation area of ​​loads 1 to N is completed in sequence and the results are cached independently, counted as buf1~bufN;

[0130] (7) ZYNQ performs DEM grid projection processing on the precise positioning solution results of multiple loads. In order to better illustrate the implementation mode and operation process of the present invention, a pixel in buf1 is taken as an example, and its precise positioning result is assumed to be (lonj, latj). i=(lonj-lon1) / 0.005, j=(latj-lat1) / 0.005 is calculated, and the pixel data is filled into the (i, j) index position of the DEM grid; the same processing is performed on all pixels in buf1 in turn to complete the DEM gridding processing of load 1. The same processing as buf1 is performed on buf2~bufN in turn to complete the DEM gridding processing of loads 2~load N. The gridding results are cached independently, counted as grid1~gridN;

[0131] (8) ZYNQ performs traversal interpolation processing on grid1 to gridN respectively, and uses bilinear interpolation to perform interpolation processing to form complete grid data grid1 to gridN;

[0132] (9) ZYNQ packages the aligned grid1 to gridN data and sends them to the processing backend for subsequent multi-payload and multi-channel joint processing. The above steps are repeated until the payload data stops transmitting or the cache data is solved.

[0133] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0134] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for real-time satellite registration of multi-payload wide-width image data, characterized in that: include: Step S1: The interface module FPGA receives the image data transmitted by multiple payloads, performs data analysis, extracts the image data and auxiliary data of multiple payloads, repackages them and sends them to ZYNQ; Step S2: Instruct ZYNQ to receive multi-payload image data and auxiliary data, and perform cache processing respectively; Step S3: When the image data cache meets the preset amount, a four-corner pixel coarse positioning operation is performed on the cached head and tail data of any load; Step S4: Calculate the maximum and minimum values ​​of longitude and latitude for the rough positioning results of the four corner point pixels, obtain the geographic coordinate range of the observation data and expand it; Step S5: extracting corresponding DEM equal-angle grid data in the external storage module according to the geographic coordinates of the observed data, and storing them in the DDR cache; Step S6: Performing coarse positioning processing on the observation data frame by frame and pixel by pixel without control points, and taking out the elevation data of the pixel point in DDR according to the latitude and longitude results of the coarse positioning, performing iterative solution of precise positioning with elevation constraints, and independently caching the precise positioning solution results of multiple loads; Step S7: Perform DEM grid projection processing on the precise positioning solution results of multiple loads respectively; Step S8: Perform traversal interpolation processing on the data after DEM grid projection of multiple loads to form complete DEM grid data; Step S9: sending the interpolated data to the processing backend for subsequent multi-load and multi-channel joint processing; Repeat steps S1 to S9 until the payload data stops transmitting or the cache data is resolved.

2. The method for real-time satellite registration of multi-payload wide-width image data according to claim 1, characterized in that: In step S2: The cache process uses a circular queue to store data, taking into account both the stability of memory usage in the onboard software and the convenience of development and use.

3. The method for real-time satellite registration of multi-payload wide-width image data according to claim 1, characterized in that: In step S4: The geographic coordinate range calculation uses the four-corner point coarse positioning method to frame the latitude and longitude grid range of the observation data, reducing the computing resource overhead caused by pixel-by-pixel positioning operations.

4. The method for real-time satellite registration of multi-payload wide-width image data according to claim 1, characterized in that: In step S5: DEM data extraction adopts the design of satellite-borne external storage module, and the communication method adopts optical fiber interface to meet the needs of DEM data storage and regional DEM data extraction; DEM data storage is achieved by using a solid-state disk plug-in. The DEM grid data resolution is 0.0005°×0.0005°. It is designed to be programmable on-orbit and can replace DEM data in any area to meet the needs of on-orbit dynamic updates.

5. The method for real-time satellite registration of multi-payload wide-width image data according to claim 1, characterized in that: In step S6: Based on the pixel precise positioning calculation constrained by DEM elevation information, the influence of elevation on positioning accuracy is comprehensively considered to make up for the defect of insufficient plane positioning accuracy without control points; The pixel precise positioning operation constrained by DEM elevation information adopts the method of pre-coarse positioning to extract DEM grid data within the observation area and store it in the DDR of the processing board, avoiding the resource overhead caused by repeated reading of the solid-state disk and improving the speed of DEM elevation data acquisition.

6. The method for real-time satellite registration of multi-payload wide-width image data according to claim 1, characterized in that: In step S7: The precise positioning solution results of multiple loads are projected onto DEM grids, and the grid data after rough projection is subjected to another traversal interpolation process to obtain complete multi-load grid data based on DEM registration; Real-time grid registration processing is performed on different types of payload data carried by satellites in orbit to solve the problem of multiple payloads observing different domains at the same time. The processed multi-payload grid data can be sent to the back end for different types of band combination interpretation processing.

7. The method for real-time satellite registration of multi-payload wide-width image data according to claim 1, characterized in that: The invention provides a satellite-borne embedded architecture of a combination of FPGA, ZYNQ and SSD storage, which adopts a preset high-throughput FPGA and a preset high-performance ARM processor, as well as a preset large-capacity and high-throughput SSD combination to meet the requirements of large real-time image throughput and high computing resources. At the same time, the SSD storage method is also used to obtain the DEM grid data of the observation area in real time and carry out pixel-by-pixel geometric precise positioning processing of multiple payloads.

8. A multi-payload wide-width image data satellite-borne real-time registration system, characterized in that: include: Module M1: The interface module FPGA receives the image data transmitted by multiple payloads, performs data analysis, extracts the image data and auxiliary data of multiple payloads, repackages them and sends them to ZYNQ; Module M2: Instructs ZYNQ to receive multi-payload image data and auxiliary data and perform cache processing respectively; Module M3: When the image data cache meets the preset amount, a four-corner pixel coarse positioning operation is performed on the cache head and tail data of any load; Module M4: Calculate the maximum and minimum values ​​of longitude and latitude for the rough positioning results of the four corner point pixels, obtain the geographic coordinate range of the observation data and expand it; Module M5: extracts the corresponding DEM equiangular grid data in the external storage module according to the geographic coordinates of the observed data, and stores it in the DDR cache; Module M6: Carry out coarse positioning processing of observation data frame by frame and pixel by pixel without control points. At the same time, according to the latitude and longitude results of coarse positioning, the elevation data of the pixel point is taken out from DDR, and the precise positioning iterative solution of elevation constraints is carried out. The precise positioning solution results of multiple loads are cached independently. Module M7: DEM grid projection processing is performed on the precise positioning solution results of multiple loads; Module M8: Perform traversal interpolation processing on the DEM grid projection data of multiple loads to form complete DEM grid data; Module M9: Packaging the interpolated data to the processing backend for subsequent multi-load and multi-channel joint processing; Modules M1 to M9 are triggered repeatedly until the payload data stops transmitting or the buffer data is resolved.

9. The multi-payload wide-width image data satellite-borne real-time registration system according to claim 8, characterized in that: In the module M2: The cache process uses a circular queue to store data, taking into account the stability of memory usage in the onboard software and the convenience of development and use; In the module M4: The geographic coordinate range calculation uses the four-corner point coarse positioning method to frame the latitude and longitude grid range of the observation data, reducing the computing resource overhead caused by pixel-by-pixel positioning operations; In the module M5: DEM data extraction adopts the design of satellite-borne external storage module, and the communication method adopts optical fiber interface to meet the needs of DEM data storage and regional DEM data extraction; DEM data storage is implemented by using a solid-state disk plug-in. The DEM grid data resolution is 0.0005°×0.0005°. The design is programmable on-orbit, and the DEM data of any area can be replaced to meet the needs of on-orbit dynamic updates. In the module M6: Based on the pixel precise positioning calculation constrained by DEM elevation information, the influence of elevation on positioning accuracy is comprehensively considered to make up for the defect of insufficient plane positioning accuracy without control points; The pixel precise positioning operation constrained by DEM elevation information adopts the method of pre-coarse positioning to extract DEM grid data within the observation area and store it in the DDR of the processing board, avoiding the resource overhead caused by repeated reading of the solid-state disk and improving the speed of DEM elevation data acquisition.

10. The multi-payload wide-width image data on-board real-time registration system according to claim 8, characterized in that: In the module M7: The precise positioning solution results of multiple loads are projected onto DEM grids, and the grid data after rough projection is subjected to another traversal interpolation process to obtain complete multi-load grid data based on DEM registration; Carry out real-time grid registration processing on different types of payload data carried by satellites in real time on orbit to solve the current situation of multiple payloads observing different domains at the same time. The processed multi-payload grid data can be sent to the back-end for different types of band combination interpretation processing; The invention provides a satellite-borne embedded architecture of a combination of FPGA, ZYNQ and SSD storage, which adopts a preset high-throughput FPGA and a preset high-performance ARM processor, as well as a preset large-capacity and high-throughput SSD combination to meet the requirements of large real-time image throughput and high computing resources. At the same time, the SSD storage method is also used to obtain the DEM grid data of the observation area in real time and carry out pixel-by-pixel geometric precise positioning processing of multiple payloads.

Images