Information processing device, information processing method, and recording medium

Abstract

This information processing device comprises a dividing unit, a difference calculation unit, a weight calculation unit, a bias calculation unit, and an arithmetic control unit. The dividing unit divides a target region into a plurality of partial regions in which adjacent partial regions partially overlap each other. The difference calculation unit calculates, for the partial regions, differences between analysis values of overlapping points corresponding in overlapping regions that are overlapping each other. The weight calculation unit calculates weights for the overlapping points included in the overlapping regions in accordance with the difference between the analysis values of the overlapping points and the difference between bias values for the partial regions including the overlapping points. The bias calculation unit uses the weights to calculate the biases values for the partial regions so as to reduce the degree of deviation between the difference in analysis values and the difference in the bias values in the whole overlapping regions. The arithmetic control unit performs control so as to repeat calculation of the weights and the bias values until a prescribed termination condition is satisfied.

Description

Information processing device, information processing method, and recording medium

[0001] This invention relates to an information processing apparatus, an information processing method, and a recording medium.

[0002] SAR images are used to detect and measure minute changes occurring on the ground, such as ground subsidence and building tilt. SAR images are obtained by observing the backscatter signals of radar waves transmitted from a SAR (Synthetic Aperture Radar) to a target area. Errors can occur in multitemporal interferometry, which analyzes wide-area SAR images, due to the effects of phase (atmospheric delay) and starting errors, which depend on the amount of water vapor in the atmosphere.

[0003] For example, Non-Patent Document 1 discloses MT-inSAR (Multi-temporal interferometric synthetic aperture radar) technology for correcting atmospheric delay and orbital errors over a wide range.

[0004] Non-patent document 1 describes how registered single look complex (SLC) data is divided and processed individually using a general MT-inSAR algorithm. Next, the results are corrected by an adjustment model based on the spatial consistency of homomy points (the same ground deformation point within overlapping regions located in different blocks). Finally, the results are combined using a weighted average method to obtain a continuous and overall deformation result. In this final combination, a block-specific constant is subtracted from the corrected result to minimize the difference in the analysis results for homomy points. The subtracted constant is primarily an error including the difference from the benchmark, and can also be called a bias.

[0005] Yuexin Wang, et al., "An MT-InSAR Data Partition Strategy for Sentinel-1A / B TOPS Data," [online], [Accessed May 13, 2024], Internet <URL: https: / / www.mdpi.com / 2072-4292 / 14 / 18 / 4562>

[0006] However, in the technology described in Non-Patent Document 1, for example, if the blocks contain outliers, the accuracy of the final analytical values ​​may be poor.

[0007] One of the challenges of this disclosure is to obtain highly accurate analytical values ​​at each point in the target region.

[0008] The information processing apparatus in this disclosure includes: division means for dividing a target area into a plurality of sub-regions in which adjacent sub-regions partially overlap each other; difference calculation means for each of the sub-regions, calculating the difference in analytical values ​​of corresponding overlapping points in the overlapping regions which are overlapping regions; weight calculation means for calculating a weight for each overlapping point according to the difference in analytical values ​​of each overlapping point and the difference in bias values ​​for the sub-regions containing each overlapping point; bias calculation means for calculating a bias value for each of the sub-regions using the weights such that the degree of deviation between the difference in analytical values ​​and the difference in bias values ​​in the entire overlapping region becomes small; and calculation control means for controlling the calculation of the weights and bias values ​​to be repeated until a predetermined termination condition is met.

[0009] The information processing method in this disclosure involves one or more computers dividing a target area into multiple sub-regions in which adjacent sub-regions partially overlap each other, calculating the difference in analysis values ​​for each overlapping point in the overlapping region (the region that overlaps with each other) for each of the sub-regions, calculating a weight for each overlapping point according to the difference in analysis values ​​for each overlapping point and the difference in bias values ​​for the sub-region containing each overlapping point, calculating a bias value for each of the sub-regions using the weight so that the degree of difference between the difference in analysis values ​​and the difference in bias values ​​in the entire overlapping region becomes small, and controlling the computer to repeat the calculation of the weight and bias value until a predetermined termination condition is met.

[0010] The recording medium in this disclosure is a recording medium on which a program is recorded that causes one or more computers to perform the following actions: divide a target area into a plurality of sub-regions in which adjacent sub-regions partially overlap each other; calculate the difference in analysis values ​​of each overlapping point in the overlapping region which is the overlapping region for each of the sub-regions; calculate a weight for each overlapping point according to the difference in analysis values ​​of each overlapping point and the difference in bias values ​​for the sub-region containing each overlapping point; use the weight to calculate a bias value for each of the sub-regions such that the degree of difference between the difference in analysis values ​​and the difference in bias values ​​in the entire overlapping region becomes small; and control the computer to repeat the calculation of the weight and the bias value until a predetermined termination condition is met.

[0011] According to this disclosure, it becomes possible to obtain highly accurate analytical values ​​at each point in the target region.

[0012] This is a block diagram showing an example configuration of the first information processing device according to this disclosure. This is a flowchart showing an example of the processing operation of the first information processing device according to this disclosure. This is a block diagram showing an example configuration of the first information processing system according to this disclosure. This is a block diagram showing a detailed example configuration of the first information processing device according to this disclosure. This is a flowchart showing a detailed example of the processing operation of the first information processing device according to this disclosure. This is a diagram showing an example in which a target area included in a radar image according to this disclosure is divided into multiple sub-regions including sub-regions i and j. This is a block diagram showing an example physical configuration of the first information processing device according to this disclosure. This is a block diagram showing an example configuration of the second information processing device according to this disclosure. This is a flowchart showing an example of the processing operation of the second information processing device according to this disclosure.

[0013] In this disclosure, the drawings are associated with one or more embodiments. In all drawings, similar components are denoted by the same reference numerals, and their descriptions are omitted where appropriate.

[0014] [Embodiment 1] (Overview) As shown in Figure 1, the information processing device 100 includes a division unit 110, a difference calculation unit 130, a weight calculation unit 140, a bias calculation unit 150, and an arithmetic control unit 160.

[0015] The division unit 110 divides the target area into multiple sub-regions in which adjacent sub-regions partially overlap each other.

[0016] The difference calculation unit 130 calculates the difference between the analytical values ​​of each overlapping point in the overlapping region, which is the region that overlaps with each other, for each of the sub-regions.

[0017] The weight calculation unit 140 calculates a weight for each overlapping point based on the difference in the analysis values ​​of each overlapping point and the difference in the bias values ​​for the sub-region containing each overlapping point.

[0018] The bias calculation unit 150 uses weights to calculate bias values ​​for each sub-region such that the degree of deviation between the difference in analysis values ​​and the difference in bias values ​​across the entire overlapping region is reduced.

[0019] The calculation control unit 160 controls the system to repeat the calculation of weights and bias values ​​until predetermined termination conditions are met.

[0020] According to this information processing device 100, when calculating the bias value, a weight corresponding to the difference between the analysis value and the bias value at overlapping points is considered. Therefore, even when a sub-region contains outliers, the bias value can be calculated while suppressing their influence. Then, the analysis value at each point in the target region can be corrected using such a bias value. Consequently, it becomes possible to obtain highly accurate analysis values ​​at each point in the target region.

[0021] The information processing device 100 performs the information processing shown in Figure 2.

[0022] The division unit 110 divides the target area into multiple sub-regions in which adjacent sub-regions partially overlap each other (step S110).

[0023] The difference calculation unit 130 calculates the difference between the analysis values ​​of each overlapping point in the overlapping region, which is the region that overlaps with the other, for each of the sub-regions (step S130).

[0024] The weight calculation unit 140 calculates the weight for each overlapping point included in the overlapping region according to the difference between the analysis values ​​of each overlapping point and the difference between the bias values ​​for the subregion including each overlapping point (step S140).

[0025] The bias calculation unit 150 uses weights to calculate bias values ​​for each sub-region such that the difference between the difference in analysis values ​​and the difference in bias values ​​across the entire overlapping region included in the target region is minimized (step S150).

[0026] The calculation control unit 160 controls the system to repeat the calculation of weights and bias values ​​(steps S140 and S150) until predetermined termination conditions are met (step S160).

[0027] According to this information processing method, when calculating the bias value, a weight is taken into account based on the difference between the analysis value and the bias value at overlapping points. Therefore, even when a sub-region contains outliers, the bias value can be calculated while suppressing their influence. Then, the analysis value at each point in the target region can be corrected using such a bias value. Consequently, it becomes possible to obtain highly accurate analysis values ​​at each point in the target region.

[0028] The following describes a detailed example of the information processing device 100.

[0029] (Detailed example) The information processing device 100 is provided, for example, in an observation system SYS for observing a target area based on the backscatter signal of radar waves. The observation system SYS comprises a flying object 10 and the information processing device 100, as shown in Figure 3, for example. Note that there may be multiple flying objects 10.

[0030] (Flying object 10) Flying object 10 irradiates the target area with radar waves and observes the backscatter signal. Flying object 10 transmits observation data based on the observed backscatter signal. Flying object 10 is, for example, a SAR (Synthetic Aperture Radar) satellite that irradiates the Earth's surface with radar waves and observes the backscatter signal. Note that flying object 10 is not limited to a SAR satellite, but may be, for example, a drone.

[0031] Observation data may include the observation time, the observation position, the method of irradiating radar waves, etc. The observation time is information representing the time when the observation was made. The observation time is, for example, a date, but the representation method is not limited to a date and may be, for example, a date and time or the like. The observation position is, for example, the position of the flying object 10 at the time of observation.

[0032] (Regarding the information processing apparatus 100) As shown in FIG. 4, for example, the information processing apparatus 100 may include an analysis unit 120 and a correction unit 170 in addition to the above-described division unit 110, difference calculation unit 130, weight calculation unit 140, bias calculation unit 150, and arithmetic control unit 160.

[0033] The analysis unit 120 calculates an analysis value of each point included in each partial region obtained by dividing the target region, for each of the partial regions obtained by dividing the target region, based on the radar image obtained by observing the target region.

[0034] The correction unit 170 calculates a corrected analysis value obtained by correcting the analysis value at each point of the target region using the bias value.

[0035] The information processing apparatus 100 executes information processing as shown in FIG. 5, for example.

[0036] The above-described step S110 is executed.

[0037] The analysis unit 120 calculates an analysis value of each point included in each partial region obtained by dividing the target region, for each of the partial regions obtained by dividing the target region, based on the radar image obtained by observing the target region (step S120).

[0038] The above-described steps S130 to S160 are executed.

[0039] The correction unit 170 calculates a corrected analysis value obtained by correcting the analysis value at each point of the target region using the bias value (step S170).

[0040] (Regarding the division unit 110) As described above, the division unit 110 divides the target region into a plurality of partial regions. Each partial region includes a region (overlapping region) where a part overlaps with an adjacent partial region.

[0041] For example, the division unit 110 may acquire a radar image obtained by observing the target area. The division unit 110 may, for example, identify the target area based on the radar image and divide the identified target area into a plurality of sub-regions.

[0042] A radar image is image information that includes, for example, observed values ​​represented by complex numerical values ​​representing amplitude and phase in the pixel value of each pixel (point), and is an interferometric SAR image, but is not limited to this.

[0043] Figure 6 shows an example of dividing a target area included in a radar image into multiple sub-regions. The sub-regions i and j shown in the figure are arbitrary sub-regions that are adjacent to each other from among the multiple sub-regions. Sub-regions i and j include an overlapping region ij that overlaps with each other.

[0044] The points included in the figure correspond to pixels. An overlapping region contains one or more overlapping points. An overlapping point is a point (e.g., a pixel) included in the overlapping region i and j, and is a corresponding point between subregions i and j, respectively. An overlapping point may also be a point included in the overlapping region of subregions i and j, and its position (e.g., pixel position) in subregions i and j is corresponding to that point. In the figure, an example is shown where the overlapping region i and j contains multiple overlapping points, one of which is shown as overlapping point k.

[0045] The size of the sub-region may be, for example, a predetermined size. The shape of the sub-region is not limited to a rectangle and may be changed as appropriate.

[0046] (Regarding the analysis unit 120) The analysis unit 120 calculates analysis values ​​for each point included in each of the sub-regions into which the target region has been divided, based on the radar image acquired by the division unit 110.

[0047] The analytical values ​​are, for example, values ​​obtained by analyzing radar images such as SAR images. The analytical values ​​may be one or more, such as displacement velocity, displacement, and elevation. The displacement velocity may be, for example, a relative velocity based on a portion of a sub-region. The following displacement velocity is an example of an analytical value.

[0048] As described above, the analytical values ​​are calculated for each sub-region. In other words, the analytical values ​​are calculated for each sub-region. Therefore, the analytical values ​​of overlapping points may differ for each sub-region containing the overlapping point due to factors such as errors in the observed values ​​included in the sub-region. For example, the analytical value of overlapping point k may differ between the analytical value of overlapping point k calculated from the observed values ​​in sub-region i and the analytical value of overlapping point k calculated from the observed values ​​in sub-region j.

[0049] The analysis unit 120 may further calculate a confidence score indicating the accuracy of the calculated analysis value for each point included in each of the sub-regions. The analysis unit 120 may further calculate a confidence score for each of the overlapping points among the points included in each of the sub-regions. That is, the analysis unit 120 may further calculate a confidence score for at least each of the overlapping points.

[0050] The confidence level may be, for example, a large value indicating a high probability of the analyzed value, a value between 0 and 1 (inclusive), or coherence.

[0051] Note that the analysis values ​​and confidence levels are not limited to those exemplified herein. Furthermore, the analysis values ​​may be calculated by an external device not shown. In this case, for example, the information processing device 100 does not need to have an analysis unit 120, and the division unit 110 may acquire analysis information, including analysis values, from the external device together with the radar image, or together with the radar image. This analysis information may further include the confidence level as described above. The division unit 110 may identify the target area based on the analysis information.

[0052] (Regarding the difference calculation unit 130) As described above, the difference calculation unit 130 calculates the difference between the analysis values ​​of each overlapping point in the overlapping region, which is the region that overlaps with each other, for each of the sub-regions.

[0053] For example, if a point k is included in the overlapping region ij, the difference in the analytical values ​​of that point k is the difference in the analytical values ​​of the corresponding point in each of the subregions i and j. In other words, the difference in the analytical values ​​of a point k is, for example, the difference between the analytical value ki of the point k calculated from the observed values ​​in subregion i and the analytical value kj of the point k calculated from the observed values ​​in subregion j. The difference in the analytical values ​​of a point k may be, for example, the value obtained by subtracting the analytical value kj from the analytical value ki, or the value obtained by subtracting the analytical value ki from the analytical value kj.

[0054] (Regarding the weight calculation unit 140) As described above, the weight calculation unit 140 calculates the weight for each overlapping point included in the overlapping region according to the difference in the analysis values ​​of each overlapping point and the difference in the bias values ​​for each of the subregions that include each overlapping point.

[0055] The weight of a point of overlap may be small (for example, close to zero) as the difference between the analytical value of the point of overlap and the difference between the bias value for the subregion containing the point of overlap deviates further. For example, the weight calculation unit 140 may calculate a smaller weight as the degree of deviation, which will be described in detail later, increases. This makes it possible to reduce the weight of points where the difference in analytical values ​​deviates significantly from the difference in bias values. When the difference in analytical values ​​deviates significantly from the difference in bias values, there is a high probability that the observed values ​​in the corresponding subregions i and j are outliers, so the weight of such points can be reduced.

[0056] The degree of deviation is a value that represents the degree of difference (deviation) between the difference in analytical values ​​and the difference in bias values ​​for overlapping points. This degree of deviation can also be said to represent the degree of difference in values ​​obtained by correcting analytical values ​​using bias values ​​for overlapping points. More specifically, for example, the degree of deviation is the absolute value of the difference between the difference in analytical values ​​at an overlapping point and the difference in bias values ​​at that overlapping point. This degree of deviation can also be said to be the absolute value of the difference in values ​​obtained by correcting analytical values ​​at an overlapping point using bias values. However, the degree of deviation is not limited to this, and may also be, for example, the ratio of values ​​obtained by correcting analytical values ​​using bias values, or the absolute value of the logarithm of said ratio. For example, the bias value may be expressed as a multiplier of the analytical value. In such cases, the absolute value of the logarithm of the ratio of values ​​obtained by correcting analytical values ​​using bias values ​​is suitable as the degree of deviation.

[0057] More specifically, for example, the weight calculation unit 140 uses the following equation (1) to calculate the weight w of the overlapping point k included in the overlapping region ij. ｉｊｋ You may calculate b. ｉ , b ｊ These are the bias values ​​of sub-regions i and j, respectively, and may be calculated, for example, using equation (2) described later before the iteration by the arithmetic control unit. Alternatively, if the iteration by the arithmetic control unit has not been performed even once, the initial values ​​of the bias values ​​initialized by the method described later may be used. v ｉ，ｋ v represents the displacement velocity v of the overlapping point k obtained for the subregion i. ｊ，ｋ This represents the displacement velocity v of the overlapping point k obtained for the subregion j.

[0058]

[0059] The function g(x) in equation (1) is, for example, a function that decreases monotonically as x increases. The function g(x) may also be, for example, a function that has a maximum value when x = 0 and approaches zero as x increases.

[0060] For example, the function g(x) is exp(-x / 2σ). ２ ) is also acceptable. With this function, g(x) is always a positive value, and even if there is a large discrepancy between the difference in analytical values ​​and the difference in bias values, the weight will not become zero, and a highly accurate correction result can be obtained.

[0061] For example, the function g(x) may be a step function that is "1" when x is less than or equal to σ and "0" when x is less than σ. In this case, if the difference between the analytical values ​​and the difference between the bias values ​​diverges significantly, the weight becomes zero, so a highly accurate correction result with excellent robustness to outliers can be obtained.

[0062] (Regarding the bias calculation unit 150) As described above, the bias calculation unit 150 calculates the bias value for each partial region so that the degree of divergence between the difference in the analysis value and the difference in the bias value in the entire overlapping region included in the target region is reduced using weights. The bias calculation unit 150 may use the weights calculated by the weight calculation unit 140 to calculate the bias value.

[0063] For example, the bias calculation unit 150 may calculate the bias value for each partial region so that the weighted sum obtained by weighting the value corresponding to the degree of divergence using the weights calculated by the weight calculation unit 140 is reduced.

[0064] Specifically, for example, the bias calculation unit 150 calculates the bias value b for each of the partial regions i and j so that the sum S represented by the following formula (2) is minimized. ｉ , b ｊ may be calculated. v ｉ，ｋ represents the displacement speed v of the overlapping point k obtained for the partial region i. v ｊ，ｋ represents the displacement speed v of the overlapping point k obtained for the partial region j.

[0065]

[0066] The function f(x) in formula (1) is, for example, a function whose value increases as x increases (i.e., a monotonically increasing function). For example, the function f(x) may be a function representing a straight line whose value increases as x increases (a straight line with a positive slope), and from the viewpoint of reducing the calculation load, for example, f(x) = x is suitable.

[0067] (Regarding the arithmetic control unit 160) As described above, the arithmetic control unit 160 controls to repeat the calculation of the weights and the bias value until a predetermined end condition is satisfied. The arithmetic control unit 160 may control, for example, so that the calculation of the weights and the bias value is sequentially repeated by each of the weight calculation unit 140 and the bias calculation unit 150.

[0068] In accordance with the control of the arithmetic control unit 160, the bias calculation unit 150 may, for example, determine the bias value calculated immediately before satisfying the end condition as the final bias value.

[0069] The predetermined termination conditions may include, for example, a predetermined number of repetitions. In this case, the calculation control unit 160 may terminate the repetitive calculation process for weights and bias values ​​when the calculation of these values ​​has been performed a predetermined number of times. The predetermined termination conditions may also include, for example, that the repeatedly calculated bias value does not change beyond a predetermined termination threshold. In this case, the calculation control unit 160 may terminate the repetitive calculation process for weights and bias values ​​when the bias value no longer changes beyond the said termination threshold.

[0070] Here, appropriate values ​​may be set for the initial values ​​of the weights and bias values. For example, predetermined constants such as 0 and 1 may be used as the initial values ​​of the weights and bias values. The initial values ​​of the weights and bias values ​​may also be different.

[0071] The initial value of the bias may be, for example, a value calculated such that the sum of the magnitudes of the differences in the corrected values ​​of the analysis values ​​of each overlapping point included in the overlapping region is small. Specifically, for example, the initial value of the bias may be, for example, the initial value of the bias of each of the subregions i and j is b ｉ０ , b ｊ０ In this case, the initial value sum T, expressed by the following formula (3), may be the value that minimizes it, or a value that is less than or equal to a predetermined value, etc.

[0072]

[0073] a in equation (3) ｉｊｋ This is an initial weight used to calculate the initial bias value for the overlapping point k, and is determined for the overlapping point k included in the overlapping region ij. ｉｊｋ This can be a predetermined constant, such as 0 or 1.

[0074] (Regarding the correction unit 180) As described above, the correction unit 180 calculates corrected analysis values ​​by correcting the analysis values ​​at each point in the target region using a bias value.

[0075] For example, the correction unit 180 calculates a corrected analysis value by correcting the analysis value of each point calculated by the analysis unit 120 for each of the target regions included in the target region using the final bias value. The correction unit 180 may also generate and output corrected analysis information including the corrected analysis value. The output method may include one or more methods such as display, storage in a memory unit, and transmission to another device.

[0076] For example, the bias value b included in the above equation (3) ｉ , b ｊ When using this method, the corrected analysis value for each point included in each subdomain is obtained by taking the analysis value for each point included in the subdomain (for example, subdomain i) and then selecting the bias value corresponding to that subdomain (for example, bias value b in the case of subdomain i). ｉ It may also be calculated by subtracting ).

[0077] Here, the corrected analysis value for each point in the overlapping region may be a value obtained by statistically processing the values ​​obtained by subtracting the bias value from the analysis values ​​of each sub-region that overlaps in the overlapping region (for example, the mean value).

[0078] (Example of physical configuration of the information processing device 100) Physically, the information processing device 100 has, for example, a bus 1010, a processor 1020, a memory 1030, a storage device 1040, a network interface 1050, an input interface 1060, and an output interface 1070, as shown in Figure 7.

[0079] Bus 1010 is a data transmission path for the processor 1020, memory 1030, storage device 1040, network interface 1050, input interface 1060, and output interface 1070 to send and receive data to and from each other. However, the method of connecting the processor 1020 and the other components to each other is not limited to bus connection.

[0080] Processor 1020 is a processor implemented using components such as a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit).

[0081] Memory 1030 is a main memory device implemented as RAM (Random Access Memory), etc.

[0082] The storage device 1040 is an auxiliary storage device implemented as an HDD (Hard Disk Drive), SSD (Solid State Drive), memory card, or ROM (Read Only Memory). The storage device 1040 stores program modules for realizing the functions of the device equipped with it. The processor 1020 reads each of these program modules into the memory 1030 and executes them, thereby realizing the functions corresponding to those program modules.

[0083] The network interface 1050 is an interface for connecting a device equipped with it to a communication network.

[0084] The input interface 1060 is an interface for the user to input information. The input interface 1060 consists of, for example, a touch panel, a keyboard, a mouse, and the like.

[0085] The output interface 1070 is an interface for presenting information to the user. The output interface 1070 is composed of, for example, a liquid crystal panel, an organic EL (Electro-Luminescence) panel, and the like.

[0086] Thus, the functions of the information processing device 100 can be realized by the collaborative execution of a software program by each physical component. Therefore, the present invention may be realized as a software program, or as a non-temporary storage medium on which the program is recorded. The information processing device may also be physically composed of multiple devices (for example, a computer).

[0087] (Function and Effects) As described above, according to this embodiment, the information processing device 100 comprises a division unit 110, a difference calculation unit 130, a first calculation unit 140, a weight calculation unit 140, a bias calculation unit 150, and an arithmetic control unit 160.

[0088] The division unit 110 divides the target region into multiple subregions in which adjacent subregions partially overlap each other. The difference calculation unit 130 calculates a first bias value for each subregion to correct the analysis value. The first calculation unit 140 calculates the difference between the analysis value and the first corrected analysis value obtained by correcting the analysis value with the first bias value for each corresponding overlapping point in the overlapping region, which is the region that overlaps in each adjacent subregion. The weight calculation unit 140 calculates a weight for each overlapping point included in the overlapping region, corresponding to the difference between the analysis value of each overlapping point and the bias value for the subregion containing each overlapping point. The bias calculation unit 150 uses the weights to calculate a bias value for each subregion such that the degree of deviation between the difference in analysis value and the difference in bias value for the entire overlapping region included in the target region becomes small. The calculation control unit 160 controls the calculation of weights and bias values ​​to be repeated until predetermined termination conditions are met.

[0089] This allows the calculation of bias values ​​to take into account a weight corresponding to the deviation between the difference in analytical values ​​and the difference in bias values. Therefore, when a sub-region contains outliers, a bias value can be calculated that suppresses their influence. Then, such bias values ​​can be used to correct the analytical values ​​at each point in the target region. Consequently, it becomes possible to obtain highly accurate analytical values ​​(corrected analytical values) at each point in the target region.

[0090] According to this embodiment, the information processing device 100 includes an analysis unit 120 that calculates analysis values ​​for each point included in each of the sub-regions into which the target region has been divided, based on a radar image obtained by observing the target region.

[0091] This allows for the correction of analysis values ​​based on radar images using bias values. Consequently, it becomes possible to obtain highly accurate analysis values ​​(corrected analysis values) at each point in the target region based on radar images.

[0092] According to this embodiment, the bias calculation unit 150 calculates a bias value for each sub-region such that the sum of weights obtained by weighting values ​​according to the degree of deviation becomes small.

[0093] This allows for the calculation of bias values ​​by considering weights corresponding to the divergence between the difference in analytical values ​​and the difference in bias values, thereby suppressing the influence of outliers in subregions. These bias values ​​can then be used to correct the analytical values ​​at each point in the target region. Consequently, it becomes possible to obtain highly accurate analytical values ​​(corrected analytical values) at each point in the target region.

[0094] According to this embodiment, the predetermined termination condition includes at least one of the following: the bias value calculated by repeating the process a predetermined number of times does not change beyond a predetermined termination threshold.

[0095] This allows for the calculation of bias values ​​to appropriately correct the analytical values, and the calculation of corrected analytical values. Therefore, it becomes possible to obtain highly accurate analytical values ​​(corrected analytical values) at each point in the target region.

[0096] (Modification 1) In Embodiment 2, an example in which the analytical value is one-dimensional was described, but the analytical value may be a vector quantity or the like that includes multiple values ​​(i.e., multiple values ​​such as displacement velocity and elevation).

[0097] In this case, equation (4) may be used instead of equation (1). Equations (5) and (6) may be used instead of equation (2).

[0098]

[0099]

[0100] Here, the bias value b ｉ , b ｊ For example, similar to equation (2), this is a bias value relating to the displacement velocity for each of the subregions i and j. Bias value d ｉ d ｊ For example, these are the bias values ​​with respect to elevation for each of the subregions i and j. The function g(x,y) in equation (4) is, for example, a function that is non-decreasing with respect to both x and y. The function g(x) is, for example, exp(-x / 2σ). ２ -y / 2σ ２) may also be the case. The function g(x) is, for example, a function that decreases monotonically as x increases. The function f(x) in equations (5) and (6) is, for example, a function whose value increases as x increases (i.e., a monotonically increasing function), similar to equation (1), and may be f(x) = x, etc.

[0101] According to this, the same effects as in Embodiment 1 are achieved with respect to multidimensional analysis values.

[0102] [Embodiment 2] Embodiment 2 describes an example of calculating the bias value using the confidence level of the analysis value. In this embodiment, in order to simplify the explanation, explanations that overlap with other embodiments will be omitted as appropriate.

[0103] (Regarding the information processing device 200) The information processing device 200 may be provided in the observation system SYS instead of the information processing device 100 described above. The information processing device 200 includes a bias calculation unit 250 instead of the bias calculation unit 150 described above, as shown in Figure 8. Except for this point, the information processing device 200 may have the same functional configuration as the information processing device 100 described above.

[0104] The bias calculation unit 250 calculates bias values ​​for each sub-region such that the sum of weights obtained by weighting values ​​according to the degree of deviation using modified weights based on the confidence level is small. The confidence level is a value that indicates the certainty of the analytical value, such as coherence as described above.

[0105] The information processing device 200 performs information processing as shown in Figure 9, for example.

[0106] Steps S110 to S150 described above are executed.

[0107] The bias calculation unit 250 calculates bias values ​​for each sub-region such that the total weighted sum, which is obtained by weighting values ​​according to the degree of deviation using modified weights based on the confidence level, becomes small.

[0108] Step S170 described above is performed.

[0109] (Regarding the bias calculation unit 250) The bias calculation unit 250 calculates a modified weight, for example, by adjusting the weight based on the confidence level. The modified weight is, for example, a value obtained by multiplying the weight by the confidence level, but is not limited to this.

[0110] The bias calculation unit 250 calculates bias values ​​for each sub-region, for example, such that the sum of weights obtained by weighting values ​​according to the degree of deviation using corrective weights becomes small. Specifically, for example, the bias calculation unit 250 calculates w in equation (2) ｉｊｋ Instead, the weighted sum S, expressed by the formula using modified weights, is minimized by the bias values ​​b for each subdomain. ｉ , b ｊ You may calculate this.

[0111] (Function and Effects) As described above, according to this embodiment, the information processing device 200 comprises an analysis unit 120 and a bias calculation unit 250. The analysis unit 120 further calculates a confidence level indicating the accuracy of the calculated analysis value for each point included in each sub-region. The bias calculation unit 250 calculates a bias value for each sub-region such that the total weighted sum, which is weighted using modified weights obtained by adjusting the weights based on the confidence level, becomes small.

[0112] This allows for the calculation of bias values ​​by considering weights corresponding to the degree of deviation. Therefore, when a sub-region contains outliers, the bias value can be calculated to suppress their influence. Furthermore, such bias values ​​can be used to correct the analytical values ​​at each point in the target region. Consequently, it becomes possible to obtain highly accurate analytical values ​​(corrected analytical values) at each point in the target region.

[0113] Although the present disclosure has been described above with reference to embodiments, the present disclosure is not limited to the embodiments described above. Various modifications to the structure and details of the present disclosure can be made as can be understood by those skilled in the art within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0114] Furthermore, while the flowcharts used in the above description show multiple steps (processes) in sequence, the execution order of the steps performed in each embodiment is not limited to the order in which they are described. In each embodiment, the order of the illustrated steps can be changed to the extent that it does not impair the content.

[0115] Some or all of the above embodiments may also be described as follows, but are not limited to the following: 1. An information processing apparatus comprising: division means for dividing a target area into a plurality of sub-regions in which adjacent sub-regions partially overlap each other; difference calculation means for each of the sub-regions, calculating the difference in analytical values ​​of corresponding overlapping points in the overlapping regions which are the overlapping regions; weight calculation means for calculating a weight for each overlapping point according to the difference in analytical values ​​of each overlapping point and the difference in bias values ​​for the sub-regions containing each overlapping point; bias calculation means for calculating a bias value for each of the sub-regions using the weights such that the degree of deviation between the difference in analytical values ​​and the difference in bias values ​​in the entire overlapping region becomes small; and calculation control means for controlling the repetition of the calculation of the weights and bias values ​​until a predetermined termination condition is met. 2. The information processing apparatus according to 1, further comprising analysis means for calculating the analytical value of each point included in each of the sub-regions into which the target area has been divided, based on a radar image obtained by observing the target area. 3. The information processing apparatus according to 1 or 2, wherein the bias calculation means calculates the bias value for each of the sub-regions such that the sum of weights obtained by weighting the value corresponding to the degree of deviation using the weights becomes small. 4. The analysis means further calculates a confidence level indicating the certainty of the calculated analysis value for each of the points included in each of the sub-regions, and the bias calculation means calculates the bias value for each of the sub-regions such that the sum of weights obtained by weighting the value corresponding to the degree of deviation using modified weights obtained by modifying the weights based on the confidence level becomes small. The information processing apparatus according to any one of 1 to 3, wherein the predetermined termination condition includes at least one of the following: the bias value calculated repeatedly for a predetermined number of times does not change beyond a predetermined termination threshold.7. An information processing method in which one or more computers divide a target region into multiple subregions in which adjacent subregions partially overlap each other, calculate the difference in analysis values ​​of corresponding overlapping points in the overlapping regions for each of the subregions, calculate a weight for each overlapping point according to the difference in analysis values ​​of each overlapping point and the difference in bias values ​​for the subregion containing each overlapping point, calculate a bias value for each of the subregions using the weight so that the degree of difference between the difference in analysis values ​​and the difference in bias values ​​in the entire overlapping region becomes small, and control the calculation of the weight and bias value to be repeated until a predetermined termination condition is met. 8. The information processing method according to 7, further comprising calculating the analysis value of each point included in each of the subregions into which the target region has been divided, based on a radar image obtained by observing the target region. 9. The information processing method according to 7. or 8., wherein the bias value for each of the sub-regions is calculated such that the sum of weights obtained by weighting the value corresponding to the degree of deviation using the weights is small. 10. The information processing method according to any one of 7. to 9., further comprising calculating a confidence level indicating the certainty of the calculated analysis value for each of the points included in each of the sub-regions, wherein the bias value for each of the sub-regions is calculated such that the sum of weights obtained by weighting the value corresponding to the degree of deviation using modified weights obtained by modifying the weights based on the confidence level is small. 11. The information processing method according to any one of 7. to 9., wherein the predetermined termination condition includes at least one of the following: the bias value calculated repeatedly for a predetermined number of times does not change beyond a predetermined termination threshold.12. A program to control one or more computers to perform the following actions: divide a target region into multiple subregions in which adjacent subregions partially overlap each other; calculate the difference in analysis values ​​of corresponding overlapping points in the overlapping regions for each of the subregions; calculate a weight for each overlapping point according to the difference in analysis values ​​of each overlapping point and the difference in bias values ​​for the subregion containing each overlapping point; use the weight to calculate a bias value for each of the subregions such that the degree of difference between the difference in analysis values ​​and the difference in bias values ​​in the entire overlapping region becomes small; and repeat the calculation of the weight and bias value until a predetermined termination condition is met. 13. The program described in 12. further, to perform the calculation of analysis values ​​for each point included in each of the subregions into which the target region has been divided, based on a radar image obtained by observing the target region. 14. The program according to 12. or 13. calculates the bias value for each of the subregions such that the sum of weights obtained by weighting the value corresponding to the degree of deviation using the weights is small. 15. The program according to any one of 12. to 14. further calculates a confidence level indicating the certainty of the calculated analysis value for each of the points included in each of the subregions, and calculates the bias value for each of the subregions such that the sum of weights obtained by weighting the value corresponding to the degree of deviation using modified weights based on the confidence level is small. 16. The program according to any one of 12. to 14. includes at least one of the predetermined termination conditions: the bias value calculated repeatedly for a predetermined number of times does not change beyond a predetermined termination threshold. 17. A recording medium on which the program according to any one of 12. to 16. is recorded.

[0116] This application claims priority based on Japanese Patent Application No. 2024-181845, filed on 17 October 2024, and incorporates all of its disclosures herein.

[0117] 100, 200 Information processing device 110 Splitting unit 120 Analysis unit 130 Difference calculation unit 140 Weight calculation unit 150, 250 Bias calculation unit 160 Calculation control unit 170 Correction unit

Claims

1. An information processing apparatus comprising: division means for dividing a target area into a plurality of sub-regions in which adjacent sub-regions partially overlap each other; difference calculation means for each of the sub-regions, calculating the difference in analytical values ​​of corresponding overlapping points in the overlapping regions which are the overlapping regions; weight calculation means for calculating a weight for each overlapping point according to the difference in analytical values ​​of each overlapping point and the difference in bias values ​​for the sub-regions containing each overlapping point; bias calculation means for calculating a bias value for each of the sub-regions using the weights such that the degree of deviation between the difference in analytical values ​​and the difference in bias values ​​in the entire overlapping region becomes small; and calculation control means for controlling the repetition of the calculation of the weights and bias values ​​until a predetermined termination condition is met.

2. The information processing apparatus according to claim 1, further comprising analysis means for calculating analysis values ​​for each point included in each of the sub-regions obtained by dividing the target region based on a radar image obtained by observing the target region.

3. The information processing apparatus according to claim 1 or 2, wherein the bias calculation means calculates the bias value for each of the subregions such that the sum of weights obtained by weighting the value corresponding to the degree of deviation using the weights becomes small.

4. The information processing apparatus according to claim 2, wherein the analysis means further calculates a confidence level indicating the certainty of the calculated analysis value for each point included in each of the sub-regions, and the bias calculation means calculates the bias value for each of the sub-regions such that the weighted sum obtained by weighting a value corresponding to the degree of deviation using a modified weight obtained by modifying the weight based on the confidence level becomes small.

5. The information processing apparatus according to claim 1 or 2, wherein the predetermined termination condition is at least one of the following: the bias value calculated by repeating the predetermined number of times does not change beyond a predetermined termination threshold.

6. An information processing method in which one or more computers divide a target area into multiple sub-regions in which adjacent sub-regions partially overlap each other, calculate the difference in analysis values ​​of corresponding overlapping points in the overlapping regions for each of the sub-regions, calculate a weight for each overlapping point according to the difference in analysis values ​​of each overlapping point and the difference in bias values ​​for the sub-region containing each overlapping point, calculate a bias value for each of the sub-regions using the weight so as to reduce the degree of difference between the difference in analysis values ​​and the difference in bias values ​​in the entire overlapping region, and control the calculation of the weights and bias values ​​to be repeated until a predetermined termination condition is met.

7. A recording medium on which a program is stored that causes one or more computers to perform the following actions: divide a target area into multiple sub-regions in which adjacent sub-regions partially overlap each other; calculate the difference in analysis values ​​of corresponding overlapping points in the overlapping regions for each of the sub-regions; calculate a weight for each overlapping point according to the difference in analysis values ​​of each overlapping point and the difference in bias values ​​for the sub-region containing each overlapping point; use the weight to calculate a bias value for each of the sub-regions such that the degree of difference between the difference in analysis values ​​and the difference in bias values ​​in the entire overlapping region becomes small; and control the system to repeat the calculation of the weights and bias values ​​until a predetermined termination condition is met.

Images