Multi-satellite remote sensing data coupling processing system and method based on virtualization technology

Abstract

The invention belongs to the technical field of satellite remote sensing image processing, and discloses a multi-satellite remote sensing data coupling processing system and method based on a virtualization technology. Creating a virtual machine based on the virtual resource pool as a computing node used by each application; fusing the multi-satellite remote sensing data by using an improved data coupling method, and carrying out image data coupling optimization on the computing node according to the satellite remote sensing data processing task; and storing the processed multi-satellite remote sensing data fusion data and the data obtained after image data coupling optimization is carried out on the computing node, and providing a channel for accessing satellite storage resources. According to the method, the correct matching between the grayed continuous image set and the corresponding brightness value continuous image set is finally promoted potentially, and the recognition efficiency is further enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of satellite remote sensing image processing, and in particular relates to a multi-satellite remote sensing data coupling processing system and method based on virtualization technology. Background Art

[0002] With the development of space vehicles and modern electronic and optical instruments, the technology of detecting and identifying long-distance vector distance research objects has been more and more widely used. In this case, the transmission and processing of data on target attitude displacement movement of various earth phenomena by remote sensing satellites using remote sensing technology and remote sensing equipment has become one of the key links in the realization of long-distance vector distance detection and recognition.

[0003] After the receiving system completes the reception and recording of the downlink satellite remote sensing data, it will be saved to media such as disks or tapes for archiving. At the same time, it will be handed over to the satellite remote sensing data processing system for cataloging and catalog injection to establish the cataloging information of the satellite remote sensing data. The satellite remote sensing data processing system will then perform radiation adjustment, geometric adjustment and other processing to generate the product data required by the user.

[0004] The satellite remote sensing data processing system is a complex system, which includes data processing unit, data management unit, task management unit and storage management unit. Due to the differences in satellite remote sensing data, in traditional satellite remote sensing data processing systems, a dedicated method is generally adopted, that is, a set of processing systems and their equipment can only be used for specific satellite remote sensing data processing.

[0005] In this way, for the processing of newly launched satellite remote sensing data, it is necessary to introduce a new satellite remote sensing data processing system, which brings the following problems:

[0006] 1) The equipment utilization rate is low and the system resources are not fully utilized;

[0007] 2) The system has poor scalability and cannot be expanded on the original system to accommodate the processing of newly released satellite remote sensing data;

[0008] 3) Multiple processing systems cannot back up each other because they are dedicated to their own use, resulting in low system availability. When a system fails, there are no replacement resources available.

[0009] 4) Multiple data processing systems are isolated from each other, making management inconvenient.

[0010] To solve the above problems, the invention patent Multi-satellite remote sensing data processing system and method based on virtualization technology (publication number CN103530168Q, publication date 20140122) discloses a multi-satellite remote sensing data processing system based on virtualization technology, including:

[0011] A virtualization management unit, used to create virtual machines based on a virtual resource pool as computing nodes used by various applications; wherein the virtual resource pool is constructed through virtualization;

[0012] A data processing unit, used to complete the cataloging processing and product production of satellite remote sensing data at each computing node; wherein the computing node includes a cataloging processing unit and a product production unit for satellite remote sensing data processing;

[0013] A task management unit, used to manage and schedule satellite remote sensing data processing tasks, and manage and schedule the computing nodes according to the satellite remote sensing data processing tasks;

[0014] The storage management unit is used to provide a channel for accessing storage resources and save the satellite remote sensing data and structured data of the storage resources.

[0015] At the same time, a multi-satellite remote sensing data processing method based on virtualization technology is disclosed, including:

[0016] Creating virtual machines based on a virtual resource pool as computing nodes used by various applications; wherein the virtual resource pool is constructed through virtualization;

[0017] The computing node includes a cataloging processing unit and a product production unit of a satellite remote sensing data processing system, which completes the cataloging processing and product production of the satellite remote sensing data;

[0018] Managing and scheduling the satellite remote sensing data, and managing and scheduling the computing nodes according to the satellite remote sensing data processing tasks;

[0019] The satellite remote sensing data and structured data are stored, and a channel for accessing satellite storage resources is provided.

[0020] Through good system architecture design, the invention can realize a multi-satellite remote sensing data processing system on a high-speed storage device and based on a virtualized environment, and the multi-satellite remote sensing data processing system can achieve high performance, high availability and high scalability, especially the expansion capability of the newly added satellite remote sensing data processing system and the expansion capability of computing nodes.

[0021] However, its technical defect is that in practical applications, its coupling capability of multi-satellite remote sensing data does not achieve good results, so the performance, availability and scalability of the obtained image information still need to be further improved. Summary of the invention

[0022] In order to overcome the problems existing in the related art, the disclosed embodiments of the present invention provide a multi-satellite remote sensing data coupling processing system and method based on virtualization technology.

[0023] The technical solution is as follows: A multi-satellite remote sensing data coupling processing method based on virtualization technology includes:

[0024] S1, creating a virtual machine based on a virtual resource pool as a computing node used by each application; wherein the virtual resource pool is constructed by virtualization, and the computing node includes a cataloging processing unit and a product production unit of a multi-satellite remote sensing data processing system, and completes the cataloging processing and product production of the multi-satellite remote sensing data;

[0025] S2, using the improved data coupling method to fuse multi-satellite remote sensing data, and performing image data coupling optimization on computing nodes according to satellite remote sensing data processing tasks;

[0026] S3, fuses the processed multi-satellite remote sensing data, saves the data after image data coupling and optimization of the computing nodes, and provides a channel for accessing satellite storage resources.

[0027] In step S2, the multi-satellite remote sensing data are fused using an improved data coupling method, including:

[0028] Step 1, respectively extracting the time domain and space domain features of the brightness continuous image set and the grayscale continuous image set in the multi-satellite remote sensing data set;

[0029] Step 2: Input the training set and parameters to train the improved data coupling model; after multiple iterations, the training obtains the data coupling pair D C and D G , mapping W and V between consecutive image sets, mapping matrix P;

[0030] Step 3: Based on the learned improved data coupling model, realistic failure representation is performed on the grayscale continuous image set and the brightness continuous image set of the test set respectively;

[0031] Step 4: Given any grayscale continuous image set in the test set, calculate the vector distance with each brightness continuous image set, and arrange them in ascending order according to the vector distance value. Select the brightness continuous image set with the smallest vector distance value as the true match of the grayscale continuous image set.

[0032] In step 2, the training set and parameters are input to train the improved data coupling model, including:

[0033] The training image samples collected by the satellite remote sensing data image capture device are a brightness value continuous image set Q, and the training image samples collected by the satellite remote sensing data image capture device are a grayscale continuous image set H;

[0034] Q=[Q 1 , Q 2 …Q N ]

[0035] H=[H 1 ,H 2 …H N ]

[0036] Where Q is the feature set of continuous brightness image set, H is the feature set of continuous grayscale image set, N is the number of image samples in the training set, Q 1 is the feature set of the first continuous brightness image set, Q 2 is the feature set of the second continuous brightness image set, Q N is the feature set of the Nth continuous brightness image set, H 1 is the feature set of the first grayscale continuous image set, H 2 is the feature set of the second grayscale continuous image set, H N is the feature set of the Nth grayscale continuous image set;

[0037] After multiple iterations, the training data coupling pair U C and U G , mapping W and V between continuous image sets, mapping matrix P, expression is:

[0038]

[0039] In the formula, O rep is the feature set between the detection targets after iterative update, Q is the feature set of the brightness continuous image set, E is the feature coefficient of the brightness image set, is a second-order iterative function, H is the feature set of the grayscale continuous image set, F is the feature coefficient of the grayscale continuous image set, W T is the mapping between asymmetric continuous image sets of brightness image samples, U C is the feature data of the brightness image set captured by satellite remote sensing data images, V T is the mapping between asymmetric continuous image sets of grayscale image samples, U G The feature data of the grayscale continuous image set captured by satellite remote sensing data images;

[0040] The mapping between asymmetric continuous image sets needs to minimize the difference between the feature sets of each detected target, expressed as:

[0041]

[0042] In the formula, O proj To minimize the difference between the feature sets of each detection target, i is the i-th feature, R is the number of features, j is the j-th feature, is the second-order iterative evaluation, is the i-th eigenvector of Q, S is the set of identical image sample pairs, is the jth eigenvector of H, μ i is the feature center of the continuous image set of the i-th detected target.

[0043] The continuous image set Q of brightness value is used as the query set, and the continuous image set H of grayscale value is used as the test set. The learned mapping matrix is ​​used to integrate the coupled mapping terms, and the expression is:

[0044]

[0045] In the formula, O mapping is the whole coupling mapping value, E is the characteristic coefficient of the brightness image set, F is the characteristic coefficient of the grayscale continuous image set, and R is the whole coupling mapping matrix between the fusion coefficients of the brightness and grayscale continuous image set features;

[0046] And determine the realistic term, which is defined as follows:

[0047]

[0048] In the formula, O disc is the realistic value, |S| is the number of detected targets,<i,j> is the set of the i-th and j-th elements, E i is the characteristic coefficient of the brightness image set at the i-th element, RF j is the integral coupling mapping value between the fusion coefficients of the bright value and the grayscale continuous image set features under the jth element, D is a set of different image sample pairs, β is the variable adjustment coefficient,<i,j> ∈S means that the i-th and j-th elements belong to the same detection target,<i,j> ∈D means that the i-th and j-th elements belong to different detection targets;

[0049] Then, the reconstruction loss of the continuous image set, the whole coupling mapping and the realism determination are analyzed through the standard objective formula, and the expression is:

[0050]

[0051] Where ρ1 is the mapping factor controlling the continuous image set, ρ2 is the mapping factor controlling the entire coupling, and α is the balance factor.

[0052] In step 3, based on the learned improved data coupling model, realistic failure representations are performed on the grayscale continuous image set and the brightness continuous image set of the test set, respectively, and the expression is:

[0053]

[0054] Where f is the realistic failure value of the grayscale continuous image set and the brightness continuous image set. is the feature data of the grayscale continuous image set captured by the T-order derivative of the satellite remote sensing data image, I is the mapping coefficient, β is the balance coefficient, and RT is the integral coupling mapping matrix of the T-order derivative between the fusion coefficients of the brightness value and grayscale continuous image set features;

[0055] Brightness image set feature data U captured based on satellite remote sensing data images C The representation coefficient g of the query continuous image set is fused, and the expression is:

[0056]

[0057] In the formula, is the feature data of the brightness image set captured by the T-order derivative of the satellite remote sensing data image, and C is the satellite remote sensing data image.

[0058] In step S2, image data coupling optimization is performed on the computing nodes according to the satellite remote sensing data processing task, including:

[0059] Step 1, performing error analysis between coupled information of computing node image data;

[0060] Step 2: Construction of error adjustment model based on bee-ant algorithm;

[0061] Step 3: Use the ant colony optimization algorithm to solve the optimal image data coupling matrix and achieve coupling adjustment.

[0062] In step 1, error analysis between coupled information of computing node image data is performed, including:

[0063] The Beidou geographic coordinate system is selected as the camera reference coordinate system of the strapdown inertial navigation system as the n system, the geocentric inertial coordinate system as the i system, the carrier coordinate system as the H system, and the DVL measurement coordinate system as the d system; the offset angle and scale factor between the H system and the d system have an impact on the accuracy of the combined camera system. is the image data coupling matrix between the H system and the d system, through The d series is photographed with a moving speed v d Converted to the shooting movement rate v under the H system b , the shooting movement rate of the DVL target posture displacement movement detection in the n system is expressed as:

[0064]

[0065] In the formula, is the shooting movement rate value of the DVL target posture displacement movement detection in the n system, is the image data coupling matrix between the n system and the b system, is the shooting movement rate of the d-frame under the DVL target posture, s is the scale factor, For installation offset angle.

[0066] In step 2, the construction of the error adjustment model based on the bee-ant algorithm includes:

[0067] The error between the SINS and DVL calculation node image data coupling information includes the scale factor and offset angle error. The error matrix is ​​as follows:

[0068]

[0069] In the formula, The shooting movement rate error between the SINS and DVL calculation node image data coupling information;

[0070] Through SINS / GPS combined photography, the reference shooting movement rate is obtained, and the expression is:

[0071]

[0072] In the formula, v b For reference shooting movement speed, The image data coupling matrix of the SINS / GPS combined camera, It is the shooting movement rate value in the n system detected by the SINS / GPS combined camera.

[0073] In step 3, the ant colony optimization algorithm is used to solve the optimal image data coupling matrix to achieve coupling adjustment, including:

[0074] According to the SINS image data coupling error model, the image data coupling vector node error δP, the shooting movement rate error δv, the satellite positioning error δε, and the shooting movement accelerometer error are selected. The expression is:

[0075]

[0076] In the formula, is the lateral tilt angle error, is the longitudinal offset angle error, is the spatial shooting angle error, is the longitude and latitude error;

[0077] The state equation of the SINS system is established as:

[0078] X SINS =F SINS +W SINS

[0079] Where, X SINS is the state equation of the SINS system, F SINS is the target posture displacement matrix, W SINS is Gaussian white noise with zero mean.

[0080] Another object of the present invention is to provide a multi-satellite remote sensing data coupling processing system based on virtualization technology, the system implements the multi-satellite remote sensing data coupling processing method based on virtualization technology, and the system comprises:

[0081] A virtual machine creation module is used to create a virtual machine based on a virtual resource pool as a computing node used by various applications; wherein the virtual resource pool is constructed through virtualization, and the computing node includes a cataloging processing unit and a product production unit of a multi-satellite remote sensing data processing system, which completes the cataloging processing and product production of the multi-satellite remote sensing data;

[0082] The data coupling module is used to fuse multi-satellite remote sensing data using an improved data coupling method and to perform image data coupling optimization on computing nodes according to satellite remote sensing data processing tasks;

[0083] The access module is used to fuse the processed multi-satellite remote sensing data, save the data after the image data coupling optimization of the computing node, and provide a channel for accessing satellite storage resources.

[0084] Combining all the above technical solutions, the beneficial effects of the present invention are as follows: the present invention has conducted a detailed study and investigation on the universality and challenges of grayscale continuous image sets in detection target recognition tasks, and proposed a detection target recognition method based on whole coupled data coupled learning based on asymmetric mapping between continuous image sets, which not only has good adaptability, but also has high-precision and automated application value for the detection target recognition scenario with grayscale continuous image sets.

[0085] The present invention can process a continuous image set of brightness values ​​and a continuous image set of grayscale values ​​simultaneously, and can avoid the self-noise of the continuous image set caused by many factors such as illumination, viewing angle, occlusion, etc., while learning data coupling and a mapping matrix. Compared with the existing methods, it can more accurately solve the cross-region recognition problem between continuous image sets of different image modes.

[0086] The present invention optimizes the relationship vector distance between continuous image sets by designing a discriminant item, making the learned data coupling and mapping more discriminative, and ultimately potentially promoting the correct matching between the grayscale continuous image set and the corresponding brightness value continuous image set, further enhancing the recognition efficiency.

[0087] The detection target recognition method proposed by the method of the present invention, which is based on the whole coupled data coupled pair learning of asymmetric mapping between continuous image sets, can achieve a higher matching rate in the application scenario where new brightness value and grayscale continuous image sets coexist, compared with the previous methods for solving the conventional detection target recognition problem. The matching rate is greatly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the principles of the present disclosure;

[0089] Figure 1 It is a flow chart of a multi-satellite remote sensing data coupling processing method based on virtualization technology provided by an embodiment of the present invention;

[0090] Figure 2 Schematic diagram of a multi-satellite remote sensing data coupling processing system based on virtualization technology provided by an embodiment of the present invention;

[0091] In the figure: 1. Virtual machine creation module; 2. Data coupling module; 3. Access module. DETAILED DESCRIPTION

[0092] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific implementation disclosed below.

[0093] Embodiment 1, as Figure 1 The multi-satellite remote sensing data processing method based on virtualization technology includes:

[0094] S1, creating a virtual machine based on a virtual resource pool as a computing node used by each application; wherein the virtual resource pool is constructed by virtualization, and the computing node includes a cataloging processing unit and a product production unit of a multi-satellite remote sensing data processing system, and completes the cataloging processing and product production of the multi-satellite remote sensing data;

[0095] S2, using the improved data coupling method to fuse multi-satellite remote sensing data, and performing image data coupling optimization on computing nodes according to satellite remote sensing data processing tasks;

[0096] S3, fuses the processed multi-satellite remote sensing data, saves the data after image data coupling and optimization of the computing nodes, and provides a channel for accessing satellite storage resources.

[0097] Embodiment 2, as a preferred embodiment, in step S2, the multi-satellite remote sensing data are fused using an improved data coupling method, including:

[0098] Step 1, respectively extracting the time domain and space domain features of the brightness continuous image set and the grayscale continuous image set in the multi-satellite remote sensing data set;

[0099] Step 2: Input the training set and parameters to train the improved data coupling model; after multiple iterations, the training obtains the data coupling pair D C and D G , mapping W and V between consecutive image sets, mapping matrix P;

[0100] Step 3: Based on the learned improved data coupling model, realistic failure representation is performed on the grayscale continuous image set and the brightness continuous image set of the test set respectively;

[0101] Step 4: Given any grayscale continuous image set in the test set, calculate the vector distance with each brightness continuous image set, and arrange them in ascending order according to the vector distance value. Select the brightness continuous image set with the smallest vector distance value as the true match of the grayscale continuous image set.

[0102] In step 2, the training set and parameters are input to train the improved data coupling model, including:

[0103] The training image samples collected by the satellite remote sensing data image capture device are a brightness value continuous image set Q, and the training image samples collected by the satellite remote sensing data image capture device are a grayscale continuous image set H;

[0104] Q=[Q 1 , Q 2 …Q N ]

[0105] H=[H 1 ,H 2 …H N ]

[0106] Where Q is the feature set of continuous brightness image set, H is the feature set of continuous grayscale image set, N is the number of image samples in the training set, Q 1 is the feature set of the first continuous brightness image set, Q 2 is the feature set of the second continuous brightness image set, Q N is the feature set of the Nth continuous brightness image set, H 1 is the feature set of the first grayscale continuous image set, H 2 is the feature set of the second grayscale continuous image set, H N is the feature set of the Nth grayscale continuous image set;

[0107] After multiple iterations, the training data coupling pair U C and U G , mapping W and V between continuous image sets, mapping matrix P, expression is:

[0108]

[0109] In the formula, O rep is the feature set between the detection targets after iterative update, Q is the feature set of the brightness continuous image set, E is the feature coefficient of the brightness image set, is a second-order iterative function, H is the feature set of the grayscale continuous image set, F is the feature coefficient of the grayscale continuous image set, W T is the mapping between asymmetric continuous image sets of brightness image samples, U C is the feature data of the brightness image set captured by satellite remote sensing data images, V T is the mapping between asymmetric continuous image sets of grayscale image samples, U G The feature data of the grayscale continuous image set captured by satellite remote sensing data images;

[0110] The mapping between asymmetric continuous image sets needs to minimize the difference between the feature sets of each detected target, expressed as:

[0111]

[0112] In the formula, O proj To minimize the difference between the feature sets of each detection target, i is the i-th feature, R is the number of features, j is the j-th feature, is the second-order iterative evaluation, is the i-th eigenvector of Q, S is the set of identical image sample pairs, is the jth eigenvector of H, μ i is the feature center of the continuous image set of the i-th detected target.

[0113] The continuous image set Q of brightness value is used as the query set, and the continuous image set H of grayscale value is used as the test set. The learned mapping matrix is ​​used to integrate the coupled mapping terms, and the expression is:

[0114]

[0115] In the formula, O mapping is the whole coupling mapping value, E is the characteristic coefficient of the brightness image set, F is the characteristic coefficient of the grayscale continuous image set, and R is the whole coupling mapping matrix between the fusion coefficients of the brightness and grayscale continuous image set features;

[0116] And determine the realistic term, which is defined as follows:

[0117]

[0118] In the formula, O disc is the realistic value, |S| is the number of detected targets,<i,j> is the set of the i-th and j-th elements, E i is the characteristic coefficient of the brightness image set at the i-th element, RFj is the integral coupling mapping value between the fusion coefficients of the bright value and the grayscale continuous image set features under the jth element, D is a set of different image sample pairs, β is the variable adjustment coefficient,<i,j> ∈S means that the i-th and j-th elements belong to the same detection target,<i,j> ∈D means that the i-th and j-th elements belong to different detection targets;

[0119] Then, the reconstruction loss of the continuous image set, the whole coupling mapping and the realism determination are analyzed through the standard objective formula, and the expression is:

[0120]

[0121] Where ρ1 is the mapping factor controlling the continuous image set, ρ2 is the mapping factor controlling the entire coupling, and α is the balance factor.

[0122] In step 3, based on the learned improved data coupling model, realistic failure representations are performed on the grayscale continuous image set and the brightness continuous image set of the test set, respectively, and the expression is:

[0123]

[0124] Where f is the realistic failure value of the grayscale continuous image set and the brightness continuous image set. is the feature data of the grayscale continuous image set captured by the T-order derivative of the satellite remote sensing data image, I is the mapping coefficient, β is the balance coefficient, and R T is the integral coupling mapping matrix of the T-order derivative between the fusion coefficients of the brightness value and grayscale continuous image set features;

[0125] Brightness image set feature data U captured based on satellite remote sensing data images C The representation coefficient g of the query continuous image set is fused, and the expression is:

[0126]

[0127] In the formula, is the feature data of the brightness image set captured by the T-order derivative of the satellite remote sensing data image, and C is the satellite remote sensing data image.

[0128] In step S2, image data coupling optimization is performed on the computing nodes according to the satellite remote sensing data processing task, including:

[0129] Step 1, performing error analysis between coupled information of computing node image data;

[0130] Step 2: Construction of error adjustment model based on bee-ant algorithm;

[0131] The bee-ant algorithm is a population-based random search method that uses the foraging behavior of bee-ant colonies as a model. The bee-ant algorithm builds a large group of individuals, called particles, where each particle represents a candidate solution. Through the simple behavior of individual particles and the information interaction within the group, the intelligent problem solving is achieved.

[0132] Each particle can be regarded as a search individual in the N-dimensional search space. The current position of the particle is a candidate solution to the corresponding optimization problem, and the flight process of the particle is the search process of the individual. The flight shooting movement rate of the particle can be dynamically adjusted according to the historical optimal position of the particle and the historical optimal position of the population. The particle has only two attributes: shooting movement rate and position. The shooting movement rate represents the speed of movement, and the position represents the direction of movement. The optimal solution searched by each particle individually is called the individual extreme value, and the optimal individual extreme value in the particle group is taken as the current global optimal solution. Iterate continuously, update the shooting movement rate and position, and finally obtain the optimal solution that meets the termination conditions.

[0133] Step 3: Use the ant colony optimization algorithm to solve the optimal image data coupling matrix and achieve coupling adjustment.

[0134] In step 1, error analysis between coupled information of computing node image data is performed, including:

[0135] The Beidou geographic coordinate system is selected as the camera reference coordinate system of the strapdown inertial navigation system as the n system, the geocentric inertial coordinate system as the i system, the carrier coordinate system as the H system, and the DVL measurement coordinate system as the d system; the offset angle and scale factor between the H system and the d system have an impact on the accuracy of the combined camera system. is the image data coupling matrix between the H system and the d system, through The d series is photographed with a moving speed v d Converted to the shooting movement rate v under the H system b , the shooting movement rate of the DVL target posture displacement movement detection in the n system is expressed as:

[0136]

[0137] In the formula, is the shooting movement rate value of the DVL target posture displacement movement detection in the n system, is the image data coupling matrix between the n system and the b system, is the shooting movement rate of the d-frame under the DVL target posture, s is the scale factor, For installation offset angle.

[0138] In step 2, the construction of the error adjustment model based on the bee-ant algorithm includes:

[0139] The error between the SINS and DVL calculation node image data coupling information includes the scale factor and offset angle error. The error matrix is ​​as follows:

[0140]

[0141] In the formula, The shooting movement rate error between the SINS and DVL calculation node image data coupling information;

[0142] Through SINS / GPS combined photography, the reference shooting movement rate is obtained, and the expression is:

[0143]

[0144] In the formula, v b For reference shooting movement speed, The image data coupling matrix of the SINS / GPS combined camera, It is the shooting movement rate value in the n system detected by the SINS / GPS combined camera.

[0145] In step 3, the ant colony optimization algorithm is used to solve the optimal image data coupling matrix to achieve coupling adjustment, including:

[0146] According to the SINS image data coupling error model, the image data coupling vector node error δP, the shooting movement rate error δv, the satellite positioning error δε, and the shooting movement accelerometer error are selected. The expression is:

[0147]

[0148] In the formula, is the lateral tilt angle error, is the longitudinal offset angle error, is the spatial shooting angle error, is the longitude and latitude error;

[0149] The state equation of the SINS system is established as:

[0150] X SINS =F SINS +W SINS

[0151] Where, X SINS is the state equation of the SINS system, F SINS is the target posture displacement matrix, W SINS is Gaussian white noise with zero mean.

[0152] Embodiment 4, as Figure 2As shown, the present invention provides a multi-satellite remote sensing data coupling processing system based on virtualization technology, the system comprising:

[0153] A virtual machine creation module 1 is used to create a virtual machine based on a virtual resource pool as a computing node used by various applications; wherein the virtual resource pool is constructed by virtualization, and the computing node includes a cataloging processing unit and a product production unit of a multi-satellite remote sensing data processing system, which completes the cataloging processing and product production of the multi-satellite remote sensing data;

[0154] Data coupling module 2, used to fuse multi-satellite remote sensing data using an improved data coupling method, and to perform image data coupling optimization on computing nodes according to satellite remote sensing data processing tasks;

[0155] Access module 3 is used to fuse the processed multi-satellite remote sensing data, save the data after the image data coupling optimization of the computing node, and provide a channel for accessing satellite storage resources.

[0156] It can be seen from the above embodiments that the SINSDVL coupling adjustment method of the present invention uses the particle swarm optimization algorithm to calibrate the computing node image data coupling information. The reference shooting movement rate and the DVL output shooting movement rate are constructed into two point sets, thereby converting the calibration problem into a WQhHQ problem, and using the bee ant colony optimization algorithm to solve the optimal image data coupling matrix. The simulation experiment and semi-physical experiment results show that the calibration method proposed in the present invention can quickly and accurately estimate the proportional factor and installation offset angle between the SINS and DVL computing node image data coupling information, so that the combined camera accuracy is greatly improved. Compared with the traditional calibration method, this algorithm does not require the QUV to perform special maneuvers or uniform motion, and can be calibrated in any motion state.

[0157] The present invention proposes a new SINS / DVL combined camera adjustment method, which introduces the bee-ant algorithm into the error estimation problem. An estimation model based on the bee-ant algorithm is established. The installation error and proportional factor error between the SINS and DVL computing node image data coupling information are used to establish the bee-ant algorithm fitness function, and the constant error matrix is ​​estimated. Compared with the traditional adjustment method, the advantages or contributions of this method can be summarized as follows: 1) This method does not require the carrier to perform special movements or uniform movements; 2) This method is simple and the coupled shooting movement rate is fast.

[0158] The installation error between SINS and DVL and the scale factor error have a great impact on the accuracy of SINS / DVL combined photography. Aiming at the problem of SINS / DVL coupling adjustment, a method using the bee-ant algorithm to estimate the scale factor error and offset angle is proposed. The results prove the effectiveness of the proposed method, and it has the following advantages over the traditional method: 1) The process is not limited to the motion state of the carrier and is suitable for shooting in any motion state; 2) The method can reliably estimate the scale factor error and offset angle; 3) The method is simple and the moving rate of coupled adjustment shooting is fast.

[0159] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0160] Since the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present invention, their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0161] Technical inspection target personnel in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in one processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment.

[0162] An embodiment of the present invention further provides a computer device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor implements the steps in any of the above-mentioned method embodiments when executing the computer program.

[0163] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0164] An embodiment of the present invention further provides an information data processing terminal, which, when executed on an electronic device, provides a user input interface to implement the steps in the above method embodiments. The information data processing terminal is not limited to mobile phones, computers, and switches.

[0165] An embodiment of the present invention further provides a server, which is used to provide a user input interface to implement the steps in the above method embodiments when executed on an electronic device.

[0166] An embodiment of the present invention provides a computer program product. When the computer program product runs on an electronic device, the electronic device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0167] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the camera device / terminal device, recording medium, computer memory, read-only memory (ROM), random access memory (RQndomQccessMemory, RQM), electrical carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0168] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technical detection target person familiar with the technical field shall make any modification, equivalent substitution and improvement within the technical scope disclosed by the present invention and within the spirit and principle of the present invention within the scope of the present invention. All such modification, equivalent substitution and improvement shall be covered within the protection scope of the present invention.

Claims

1. A multi-satellite remote sensing data coupling processing method based on virtualization technology, It is characterized in that The method includes: S1, creating a virtual machine based on a virtual resource pool as a computing node used by each application; wherein the virtual resource pool is constructed by virtualization, and the computing node includes a cataloging processing unit and a product production unit of a multi-satellite remote sensing data processing system, and completes the cataloging processing and product production of the multi-satellite remote sensing data; S2, using the improved data coupling method to fuse multi-satellite remote sensing data, and performing image data coupling optimization on computing nodes according to satellite remote sensing data processing tasks; S3, fuses the processed multi-satellite remote sensing data, saves the data after image data coupling and optimization of the computing nodes, and provides a channel for accessing satellite storage resources.

2. According to the multi-satellite remote sensing data coupling processing method based on virtualization technology according to claim 1, It is characterized in that In step S2, the multi-satellite remote sensing data are fused using an improved data coupling method, including: Step 1, respectively extracting the time domain and space domain features of the brightness continuous image set and the grayscale continuous image set in the multi-satellite remote sensing data set; Step 2: Input the training set and parameters to train the improved data coupling model; after multiple iterations, the training obtains the data coupling pair D C and D G , mapping W and V between consecutive image sets, mapping matrix P; Step 3: Based on the learned improved data coupling model, realistic failure representation is performed on the grayscale continuous image set and the brightness continuous image set of the test set respectively; Step 4: Given any grayscale continuous image set in the test set, calculate the vector distance with each brightness continuous image set, and arrange them in ascending order according to the vector distance value. Select the brightness continuous image set with the smallest vector distance value as the true match of the grayscale continuous image set.

3. According to the multi-satellite remote sensing data coupling processing method based on virtualization technology as described in claim 2, It is characterized in that In step 2, the training set and parameters are input to train the improved data coupling model, including: The training image samples collected by the satellite remote sensing data image capture device are the brightness value continuous image set Q, and the training image samples collected by the satellite remote sensing data image capture device are the grayscale continuous image set H; Q=[Q 1 ,Q 2 …Q N ] H=[H 1 ,H 2 …H N ] Where Q is the feature set of continuous brightness image set, H is the feature set of continuous grayscale image set, N is the number of image samples in the training set, Q 1 is the feature set of the first continuous brightness image set, Q 2 is the feature set of the second continuous brightness image set, Q N is the feature set of the Nth brightness continuous image set, H 1 is the feature set of the first grayscale continuous image set, H 2 is the feature set of the second grayscale continuous image set, H N is the feature set of the Nth grayscale continuous image set; After multiple iterations, the training data coupling pair U C and U G , mapping W and V between continuous image sets, mapping matrix P, expression is: In the formula, O rep is the feature set between the detection targets after iterative update, Q is the feature set of the brightness continuous image set, E is the feature coefficient of the brightness image set, is a second-order iterative function, H is the feature set of the grayscale continuous image set, F is the feature coefficient of the grayscale continuous image set, W T is the mapping between asymmetric continuous image sets of brightness image samples, U C is the feature data of the brightness image set captured by satellite remote sensing data images, V T is the mapping between asymmetric continuous image sets of grayscale image samples, U G The feature data of the grayscale continuous image set captured by satellite remote sensing data images; The mapping between asymmetric continuous image sets needs to minimize the difference between the feature sets of each detected target, expressed as: In the formula, O proj To minimize the difference between the feature sets of each detection target, i is the i-th feature, R is the number of features, j is the H-th feature, is the second-order iterative evaluation, is the i-th eigenvector of Q, S is the set of identical image sample pairs, is the jth eigenvector of H, μ i is the feature center of the continuous image set of the i-th detected target.

4. According to the multi-satellite remote sensing data coupling processing method based on virtualization technology as described in claim 3, It is characterized in that The continuous image set Q of brightness value is used as the query set, and the continuous image set H of grayscale value is used as the test set. The learned mapping matrix is ​​used to integrate the coupled mapping terms, and the expression is: In the formula, O mapping is the whole coupling mapping value, E is the characteristic coefficient of the brightness image set, F is the characteristic coefficient of the grayscale continuous image set, and R is the whole coupling mapping matrix between the fusion coefficients of the brightness and grayscale continuous image set features; And determine the realistic term, which is defined as follows: In the formula, O disc is the realistic value, |S| is the number of detected targets,<i,j> is the set of the i-th and j-th elements, E i is the characteristic coefficient of the brightness image set at the i-th element, RF j is the integral coupling mapping value between the fusion coefficients of the bright value and the grayscale continuous image set features under the jth element, D is a set of different image sample pairs, β is the variable adjustment coefficient,<i,j> ∈S means that the i-th and j-th elements belong to the same detection target,<i,j> ∈D means that the i-th and j-th elements belong to different detection targets; Then, the reconstruction loss of the continuous image set, the whole coupling mapping and the realism determination are analyzed through the standard objective formula, and the expression is: In the formula, ρ 1 To control the mapping factor within a continuous image set, ρ 2 is the control factor for the whole coupling mapping, and α is the balance factor.

5. According to the multi-satellite remote sensing data coupling processing method based on virtualization technology according to claim 4, It is characterized in that In step 3, based on the learned improved data coupling model, realistic failure representations are performed on the grayscale continuous image set and the brightness continuous image set of the test set, respectively, and the expression is: Where f is the realistic failure value of the grayscale continuous image set and the brightness continuous image set. is the feature data of the grayscale continuous image set captured by the T-order derivative of the satellite remote sensing data image, I is the mapping coefficient, β is the balance coefficient, and R T is the integral coupling mapping matrix of the T-order derivative between the fusion coefficients of the brightness value and grayscale continuous image set features; Brightness value image set feature data U captured based on satellite remote sensing data image C The representation coefficient g of the query continuous image set is fused, and the expression is: In the formula, is the feature data of the brightness image set captured by the T-order derivative of the satellite remote sensing data image, and C is the satellite remote sensing data image.

6. The multi-satellite remote sensing data coupling processing method based on virtualization technology according to claim 2, It is characterized in that In step S2, image data coupling optimization is performed on the computing nodes according to the satellite remote sensing data processing task, including: Step 1, performing error analysis between coupled information of computing node image data; Step 2: Construction of error adjustment model based on bee-ant algorithm; Step 3: Use the ant colony optimization algorithm to solve the optimal image data coupling matrix and achieve coupling adjustment.

7. The multi-satellite remote sensing data coupling processing method based on virtualization technology according to claim 6, It is characterized in that In step 1, error analysis between coupled information of computing node image data is performed, including: The Beidou geographic coordinate system is selected as the camera reference coordinate system of the strapdown inertial navigation system as the n system, the geocentric inertial coordinate system as the i system, the carrier coordinate system as the H system, and the DVL measurement coordinate system as the d system; the offset angle and scale factor between the H system and the d system have an impact on the accuracy of the combined camera system. is the image data coupling matrix between the H system and the d system, through The d series is photographed with a moving speed v d Converted to the shooting movement rate v under the H system b , the shooting movement rate of the DVL target posture displacement movement detection in the n system is expressed as: In the formula, is the shooting movement rate value of the DVL target posture displacement movement detection in the n system, is the image data coupling matrix between the n system and the b system, is the shooting movement rate of the d-frame under the DVL target posture, s is the scale factor, For installation offset angle.

8. The multi-satellite remote sensing data coupling processing method based on virtualization technology according to claim 7, It is characterized in that In step 2, the construction of the error adjustment model based on the bee-ant algorithm includes: The error between the SINS and DVL calculation node image data coupling information includes the scale factor and offset angle error. The error matrix is ​​as follows: In the formula, The shooting movement rate error between the SINS and DVL calculation node image data coupling information; Through SINS / GPS combined photography, the reference shooting movement rate is obtained, and the expression is: In the formula, v b For reference shooting movement speed, The image data coupling matrix of the SINS / GPS combined camera, It is the shooting movement speed value in the n system detected by the SINS / GPS combined camera.

9. The multi-satellite remote sensing data coupling processing method based on virtualization technology according to claim 8, It is characterized in that In step 3, the ant colony optimization algorithm is used to solve the optimal image data coupling matrix to achieve coupling adjustment, including: According to the SINS image data coupling error model, the image data coupling vector node error δP, the shooting movement rate error δv, the satellite positioning error δε, and the shooting movement accelerometer error are selected. The expression is: In the formula, is the lateral tilt angle error, is the longitudinal offset angle error, is the spatial shooting angle error, is the longitude and latitude error; The state equation of the SINS system is established as: X SINS =F SINS +W SINS Where, X SINS is the state equation of the SINS system, F SINS is the target posture displacement matrix, W SINS is Gaussian white noise with zero mean.

10. A multi-satellite remote sensing data coupling processing system based on virtualization technology, It is characterized in that The system implements the multi-satellite remote sensing data coupling processing method based on virtualization technology as described in any one of claims 1 to 9, and the system includes: A virtual machine creation module (1) is used to create a virtual machine based on a virtual resource pool as a computing node used by various applications; wherein the virtual resource pool is constructed by virtualization, and the computing node includes a catalog processing unit and a product production unit of a multi-satellite remote sensing data processing system, and completes the catalog processing and product production of the multi-satellite remote sensing data; A data coupling module (2) is used to fuse multi-satellite remote sensing data using an improved data coupling method, and to perform image data coupling optimization on computing nodes according to satellite remote sensing data processing tasks; The access module (3) is used to fuse the processed multi-satellite remote sensing data, save the data after the image data coupling optimization of the computing node, and provide a channel for accessing satellite storage resources.

Images