Signal processing device and signal processing method

Abstract

This signal processing device 10 comprises: a reliable three-dimensional information reconstruction unit 11 that generates reliable three-dimensional information including three-dimensional information consisting of estimated values of reflection intensity and phase at three-dimensional positions in a steady state reconstructed using observed SAR images and information indicating the reliability of the three-dimensional information; and a simulated SAR image generation unit 12 that generates a simulated SAR image that is a complex image showing a steady state suitable for imaging conditions of the SAR image to be analyzed, by using the three-dimensional information and the imaging conditions of the SAR image to be analyzed, and calculates reliability information that represents the reliability of the simulated SAR image.

Description

Signal processing device and signal processing method

[0001] The present invention relates to a signal processing device and a signal processing method using SAR images.

[0002] Patent Documents 1 to 4 describe change detection techniques using SAR (Synthetic Aperture Radar) images.

[0003] European Patent Application Publication No. 3540462, British Patent Application Publication No. 2553284, U.S. Patent No. 10571560, International Publication No. 2008 / 016153

[0004] However, Patent Documents 1 to 4 do not describe a technique for detecting a change from a steady state in a SAR image to be analyzed using a complex correlation coefficient. The steady state is represented, for example, by a predetermined complex image. Note that the photographing conditions of the complex image representing the steady state are consistent with the photographing conditions of the SAR image to be analyzed. Furthermore, for example, the disaster countermeasure support method described in Patent Document 4 uses only the reflection intensity of the reflected wave. In other words, the techniques described in Patent Documents 1 to 4 cannot correctly detect a change from a steady state.

[0005] In this specification, a state in which there is no change in the object of observation, or a state in which even if there is a change in the object of observation, the change cannot be detected, is referred to as a "steady state."

[0006] An object of the present invention is to provide a signal processing device and a signal processing method that can correctly detect changes in an object.

[0007] The signal processing device according to the present invention includes a reliable three-dimensional information reconstruction means for generating reliable three-dimensional information including three-dimensional information composed of estimated values ​​of reflection intensity and phase at three-dimensional positions in a steady state reconstructed using an observed SAR image and information indicating the reliability of the three-dimensional information, and a simulated SAR image generation means for generating a simulated SAR image, which is a complex image indicating a steady state appropriate for the imaging conditions of the SAR image to be analyzed, using the three-dimensional information and the imaging conditions of the SAR image to be analyzed, and for calculating reliability information indicating the reliability of the simulated SAR image.

[0008] The signal processing method according to the present invention is a method in which a computer generates three-dimensional information composed of estimated values ​​of reflection intensity and phase at three-dimensional positions in a steady state reconstructed using observed SAR images, and reliable three-dimensional information including information indicating the reliability of the three-dimensional information, and generates a simulated SAR image, which is a complex image indicating a steady state appropriate for the imaging conditions of the SAR image to be analyzed, using the three-dimensional information and the imaging conditions of the SAR image to be analyzed, and calculates reliability information indicating the reliability of the simulated SAR image.

[0009] The signal processing program according to the present invention causes a computer to generate three-dimensional information composed of estimated values ​​of reflection intensity and phase at three-dimensional positions in a steady state reconstructed using observed SAR images, and reliable three-dimensional information including information indicating the reliability of the three-dimensional information, and to generate a simulated SAR image, which is a complex image indicating a steady state appropriate for the imaging conditions of the SAR image to be analyzed, using the three-dimensional information and the imaging conditions of the SAR image to be analyzed, and to calculate reliability information indicating the reliability of the simulated SAR image.

[0010] According to the present invention, it is possible to correctly detect changes in an object.

[0011] 1 is an explanatory diagram showing the relationship between a coherence value and the correlation between reflection intensity and phase between SAR images. FIG. 1 is an explanatory diagram showing an example of a SAR satellite photographing a building. FIG. 2 is an explanatory diagram showing an example of a layover phenomenon occurring in a SAR image. FIG. 2 is a block diagram showing an example of the configuration of a signal processing device of a reference example. FIG. 3 is an explanatory diagram showing an example of a method for estimating a complex reflection intensity distribution at a pixel corresponding to an azimuth-range position by SAR tomography. FIG. 4 is a flowchart showing signal processing executed by a signal processing device of a reference example. FIG. 5 is a flowchart showing 3D information reconstruction processing in a reference example. FIG. 6 is a flowchart showing simulated SAR image generation processing in a reference example. FIG. 7 is an explanatory diagram for explaining change detection processing when information other than the steady state is included in a SAR image showing a steady state. FIG. 8 is a block diagram showing an example of the configuration of a signal processing device of a first embodiment. FIG. 9 is an explanatory diagram for explaining improved change detection processing when information other than the steady state is included in an observed SAR image showing a steady state. FIG. 10 is a block diagram showing another example of the configuration of the signal processing device of the first embodiment. FIG. 11 is an explanatory diagram for explaining another improved change detection processing when information other than the steady state is included in an observed SAR image showing a steady state. FIG. 12 is a flowchart showing signal processing executed by the signal processing device of the first embodiment. 1 is a flowchart showing three-dimensional information reconstruction processing in the first embodiment. FIG. 2 is a flowchart showing simulated SAR image generation processing in the first embodiment. FIG. 3 is a flowchart showing signal processing performed by a signal processing device of another aspect of the first embodiment. FIG. 4 is a flowchart showing three-dimensional information reconstruction processing in another aspect of the first embodiment. FIG. 5 is a flowchart showing simulated SAR image generation processing in another aspect of the first embodiment. FIG. 6 is a block diagram showing an example of the configuration of a signal processing device of a second embodiment. FIG. 7 is an explanatory diagram showing a specific example of first change detection processing performed by a first change detection unit. FIG. 8 is a flowchart showing signal processing performed by the signal processing device of the second embodiment. FIG. 9 is a flowchart showing first change detection processing performed by the first change detection unit. FIG. 10 is a block diagram showing an example of the configuration of a signal processing device of a third embodiment. FIG. 11 is a flowchart showing signal processing performed by the signal processing device of the third embodiment. FIG. 12 is a flowchart showing second change detection processing performed by a second change detection unit.1 is a block diagram showing an example of a computer having a CPU; FIG. 2 is a block diagram showing the main parts of a signal processing device;

[0012] Background subtraction detection technology and anomaly detection technology are known in image analysis. These technologies are used, for example, to detect objects other than permanently present objects such as buildings. These technologies are also used to detect ground conditions such as the construction or collapse of buildings. Note that objects other than permanently present objects include, for example, vehicles and aircraft. However, vehicles and aircraft can also be considered permanently present objects if they exist throughout the monitoring period.

[0013] The above technique detects changes in the image being analyzed relative to image data representing a steady state. The image data representing the steady state is selected from previously acquired and stored image data. Alternatively, the image data representing the steady state may be generated from the stored image data.

[0014] A SAR image is a complex image that contains information on the reflection intensity and phase of irradiated microwaves for each pixel. One of the change detection techniques for complex images is coherent change detection technology.

[0015] Coherent change detection technology is a technique that detects subtle changes from a steady state based on the value of the complex correlation coefficient (coherence), which indicates the similarity between images. Coherence, which is the complex correlation coefficient between SAR images, is an index that indicates the correlation between intensity and phase in a local region between SAR images. Techniques that use coherence to detect changes can increase the sensitivity of change detection because they use phase information in addition to intensity.

[0016] Figure 1 is an explanatory diagram showing the relationship between the coherence value and the correlation of the reflection intensity and phase between SAR images. As shown in Figure 1, the larger the coherence value, the higher the similarity of the reflection intensity between SAR images and the higher the similarity of the phase between SAR images. In other words, the larger the coherence value, the less variation there is between SAR images.

[0017] Also, as shown in Figure 1, the smaller the coherence value, the lower the similarity in intensity and / or phase between the SAR images, i.e., the smaller the coherence value, the more variation there is between the SAR images.

[0018] Furthermore, the coherence value is an index that is more sensitive to changes in the phase of the reflected waves than to changes in the reflection intensity. As shown in Figure 1, the coherence value becomes lower when the phase similarity is lower than the reflection intensity similarity. Coherent change detection technology can also detect changes in reflectors.

[0019] One challenge is to generate a complex image that exhibits steady state conditions suitable for SAR image analysis.

[0020] SAR images observed by SAR satellites orbiting the Earth are captured under different imaging conditions for each observation. The imaging conditions include the position of the SAR satellite at the time of observation, the coordinates of the area to be analyzed, resolution, and the like. Therefore, even when a steady-state area is observed, differences occur between the SAR images obtained for each observation. In particular, when an area with buildings varying in elevation, such as an urban area, is observed, phase information changes significantly for each observation. In such a situation, if a complex image with consistent imaging conditions (a complex image showing a steady state) cannot be generated, changes in the image due to differences in imaging conditions will be interpreted as changes occurring within the area to be analyzed. In other words, changes in the image due to differences in imaging conditions can lead to false detection. The SAR satellite is an artificial satellite equipped with a SAR.

[0021] Even when a target object is observed in a steady state, the height of the structure is one of the reasons why the observed SAR image differs depending on the shooting conditions. Figure 2 is an explanatory diagram showing an example of a SAR satellite shooting a building.

[0022] The imaging conditions for Observation 1 and Observation 2 shown in Figure 2 are different. As a result, the amount of phase change depending on the height of the building in Observation 1 is different from the amount of phase change depending on the height of the building in Observation 2. In other words, in this case, even when an observation target in a steady state is observed, the phases of the observed SAR images are different.

[0023] Even when a steady-state observation target is observed, the observed SAR image may differ depending on the shooting conditions due to the reflection intensity of overlapping buildings. Figure 3 is an explanatory diagram showing an example of the layover phenomenon that occurs in SAR images.

[0024] 3, when distance SA<distance SB, the position A of the building and the position B of the house are reversed on the image. This position reversal phenomenon is the layover phenomenon.

[0025] When a SAR satellite photographs an area including buildings and houses under conditions where a layover phenomenon occurs, a two-dimensional image (SAR image) is captured in which the buildings and houses overlap, as shown in Figure 3. The area in the two-dimensional image in Figure 3 where the buildings and houses overlap is a layover area where signals received from multiple reflectors overlap. Because the information from multiple buildings overlaps, it is difficult to extract information about each individual building from the layover area.

[0026] In layover areas, the phase of the SAR image obtained for each observation also depends on the reflection intensity of overlapping buildings. Since the overlapping state of buildings varies depending on the shooting conditions, the phase of the observed SAR image will differ even if the steady state of the observation target remains unchanged.

[0027] Due to the above two reasons, even when a SAR satellite observes a steady-state area, the SAR image obtained varies from observation to observation. In other words, the amplitude and phase of pixels that constitute permanently existing objects such as buildings in a SAR image depend on the conditions under which the SAR image was taken.

[0028] Due to the fact that the SAR image of the steady-state region obtained for each observation is different, it is difficult to estimate a complex image showing the steady state that matches the capture conditions of the SAR image to be analyzed.

[0029] If it is not possible to estimate a complex image with consistent imaging conditions (a complex image showing a steady state), it is difficult to detect changes from the complex image of the SAR image being analyzed using coherent change detection technology that uses phase information. In particular, in areas with many tall man-made structures, such as urban areas, layover phenomena specific to SAR images are likely to occur. This makes it even more difficult to estimate a complex image showing a steady state with consistent imaging conditions.

[0030] As a method for obtaining correlation using phase information other than the coherence value, for example, there is a method for extracting phase information from a SAR image as real values ​​and obtaining correlation using the extracted real values.

[0031] Next, a reference example that corresponds to the premise of the present invention will be described below. Fig. 4 is a block diagram showing an example of the configuration of a signal processing device as the reference example.

[0032] The signal processing device 500 shown in FIG. 4 generates a simulated SAR image from a plurality of observed SAR images stored in a SAR image storage unit 600 .

[0033] The simulated SAR image refers to a complex image (a complex image showing a steady state) suitable for the imaging conditions of the SAR image to be analyzed. Specifically, the simulated SAR image is a two-dimensional image in which the three-dimensional information of the reconstructed region to be analyzed is simulated using reflection intensity information and phase information. Here, the reflection intensity and phase are the reflection intensity and phase predicted to be observed when the image is captured under the same imaging conditions as the SAR image to be analyzed. In other words, a "complex image suitable for the imaging conditions of the SAR image to be analyzed (a complex image showing a steady state)" refers to a complex image that can be considered to have been captured under the same imaging conditions as the SAR image to be analyzed, i.e., the observed SAR image. The three-dimensional information is represented by data having information on reflection intensity and phase at three-dimensional positions in a steady state.

[0034] The SAR image to be analyzed is a SAR image to which motion detection is applied. The SAR image to be analyzed is obtained by photographing it with a SAR satellite. Therefore, the simulated SAR image is a complex image that represents a steady state suitable for the photographing conditions of the SAR image to be analyzed.

[0035] 4, the signal processing device 500 includes a three-dimensional information reconstruction unit 510 and a simulated SAR image generation unit 520. Also, as shown in FIG. 4, the signal processing device 500 inputs an observed SAR image from a SAR image storage unit 600.

[0036] When the three-dimensional information reconstruction unit 510 is included in a device other than the signal processing device 500, the signal processing device 500 includes only the simulated SAR image generation unit 520.

[0037] The SAR image storage unit 600 stores a plurality of observed SAR images. The SAR image storage unit 600 may be included in the signal processing device 500.

[0038] A plurality of observation SAR images stored in the SAR image storage unit 600 are input to the three-dimensional information reconstruction unit 510. The input observation SAR image is a complex image having, for each pixel, information on the reflection intensity and phase of the irradiated microwaves. The observation SAR image also includes information on the imaging conditions, such as the position of the SAR satellite at the time of observation, the coordinates of the area to be analyzed, and the resolution. Note that the three-dimensional information reconstruction unit 510 is not limited to observation SAR images obtained by observing a steady state, and may also input observation SAR images obtained when there is a change from the steady state in the area to be analyzed.

[0039] The three-dimensional information reconstruction unit 510 has a function of reconstructing and outputting three-dimensional information of the region to be analyzed (data having reflection intensity information and phase information at each three-dimensional position in a steady state). The three-dimensional information output by the three-dimensional information reconstruction unit 510 may be a three-dimensional complex reflection intensity distribution having reflection intensity information and phase information. The three-dimensional information may also be three-dimensional point cloud data, which is a set of points having reflection intensity information and phase information. Note that the three-dimensional information may include information such as temperature and displacement in addition to reflection intensity information and phase information.

[0040] SAR tomography is a method for reconstructing three-dimensional information. SAR tomography is a technique for estimating the complex reflection intensity distribution in the elevation direction for each pixel using multiple observed SAR images. The elevation direction can be defined as the direction perpendicular to the azimuth-range plane (the plane formed by the SAR satellite's direction of travel and line of sight).

[0041] That is, the three-dimensional information is information about each point in a three-dimensional space having an azimuth direction, a range direction, and an elevation direction. Each point has information about reflection intensity (an estimated value of reflection intensity) and information about phase (an estimated value of phase).

[0042] Figure 5 shows the azimuth-range position (x) obtained by SAR tomography. → 10 is an explanatory diagram showing an example of a method for estimating the complex reflection intensity distribution at a pixel corresponding to

[0043] In this specification, the symbol "→" used in the text should be written directly above the character immediately preceding it, but due to limitations on notation, it is written immediately after the character in question.

[0044] In FIG. 5, s represents the elevation direction. In FIG. 5, the plane perpendicular to the direction s represents the azimuth-range plane. The azimuth-range position (x → ) is the intersection of the axis indicating the elevation direction and the axis indicating the line of sight of the satellite in FIG.

[0045] The three-dimensional information reconstruction unit 510 estimates, for each pixel, a complex reflection intensity distribution that indicates the height, reflection intensity, and phase of buildings that are constantly present over the observation period, based on a plurality of observation SAR images.

[0046] When SAR tomography is used in the three-dimensional information reconstruction unit 510, a three-dimensional complex reflection intensity distribution is generated by combining the complex reflection intensities obtained for each pixel for all pixels in the area to be analyzed.

[0047] In order to reconstruct a three-dimensional steady state using SAR tomography or the like, a plurality of observation SAR images taken from slightly different orbits are used, and therefore, a plurality of observation SAR images are input to the three-dimensional information reconstruction unit 510.

[0048] 5 shows the first to Nth observations by the SAR satellite, where N represents the total number of observations and is an integer greater than 1.

[0049] The first to Nth observations shown in FIG. 5 correspond to a synthetic aperture in the elevation direction. In the nth (1≦n≦N) observation, the azimuth-range position (x → The relational expression between the received signal (complex signal) recorded in the pixel corresponding to (a pixel corresponding to) and the complex reflection intensity distribution at that pixel is expressed by, for example, the following equation (1):

[0050]

[0051] g in Equation (1) obs (x → ,n) is the azimuth-range position (x → ) represents the received signal (complex signal) recorded at the pixel corresponding to the → (x → ,n) is the azimuth-range position (x → ) represents the steering vector at the pixel corresponding to the α → (x → ,n) is the azimuth-range position (x → ) represents the complex reflection intensity distribution at the pixel corresponding to

[0052] The steering vector is obtained from the imaging conditions. For example, the steering vector r → (x → , n) is expressed by the following equation (2).

[0053]

[0054] k in Equation (2) n represents the altitude-phase conversion coefficient (coefficient for converting between phase and altitude). l(l = 1, ..., L) represents the position in the elevation direction. The steering vector may be expressed by an equation other than equation (2). For example, a steering vector that takes into account the effects of temperature and displacement may be used. j in equation (2) represents the imaginary unit. π in equation (2) represents the constant pi. Exp represents an exponential function with Napier's constant as the base. C represents a complex number.

[0055] A relational expression such as Equation (1) is determined for each of the first to Nth observations. The three-dimensional information reconstruction unit 510 solves an optimization problem using multiple pieces of observation data (received signals) and steering vectors to obtain the azimuth-range position (x → ) the complex reflection intensity distribution α of a permanently existing building at the pixel corresponding to bg → (x → ) is obtained.

[0056] That is, the three-dimensional information reconstruction unit 510 determines the complex reflection intensity distribution so that the N observation data of the area including the building and the like matches the N steering vectors, thereby obtaining the complex reflection intensity distribution α of the building. bg → (x → ) is obtained.

[0057] The bottom part of FIG. 5 shows the azimuth-range position (x → ) the complex reflection intensity distribution α of a permanently existing building at the pixel corresponding to bg → (x → ) absolute value|α bg The example of | is shown. Absolute value |α bg corresponds to the reflection intensity on the vertical axis shown in the lower part of Fig. 5. The horizontal axis shown in the lower part of Fig. 5 indicates the elevation position.

[0058] The complex reflection intensity distribution α shown in FIG. bg → (x → ) is the elevation position s l1 (ground), s l2 (house), s l3(building) shows a large value, i.e., position x → The received signal at elevation position s l1 , s l2 , s l3 More precisely, the signal is the overlap of the complex reflection intensities at position x → The received signal at corresponds to the result of Fourier transform of the complex reflection intensity distribution in the elevation direction.

[0059] When SAR tomography is used in the three-dimensional information reconstruction unit 510, the three-dimensional information reconstruction unit 510 creates a three-dimensional complex reflection intensity distribution by combining the complex reflection intensity distribution for each pixel for all pixels in the area to be analyzed.The three-dimensional information reconstruction unit 510 then outputs the three-dimensional complex reflection intensity distribution as three-dimensional information.Instead of creating a three-dimensional complex reflection intensity distribution, the three-dimensional information reconstruction unit 510 may create, as the three-dimensional information, three-dimensional point cloud data, which is a set of points having information on the position, reflection intensity, and phase of a reflector.

[0060] The simulated SAR image generating unit 520 receives the three-dimensional information reconstructed by the three-dimensional information reconstructing unit 510 .

[0061] The simulated SAR image generator 520 also receives input of the shooting conditions of one or more SAR images to be analyzed. The SAR images to be analyzed are, for example, one or more observed SAR images selected from the plurality of observed SAR images stored in the SAR image storage unit 600. Hereinafter, the plurality of observed SAR images may be referred to as a group of observed SAR images. In other words, the group of observed SAR images includes a plurality of observed SAR images.

[0062] The SAR image to be analyzed may be one or more newly acquired observation SAR images, or the multiple SAR images to be analyzed may be a mixture of observation SAR images stored in the SAR image storage unit 600 and newly acquired observation SAR images.

[0063] The shooting conditions of the SAR image to be analyzed that are input to the simulated SAR image generation unit 520 are, for example, the shooting conditions of the observation SAR image used to reconstruct the three-dimensional information in the three-dimensional information reconstruction unit 510. However, they may also be the shooting conditions of one or more newly acquired observation SAR images. The shooting conditions include the position of the SAR satellite at the time of observation, the coordinates of the area to be analyzed, the resolution, etc. The shooting conditions may be the shooting conditions of the observation SAR image stored in the SAR image storage unit 600, or the shooting conditions of a newly acquired observation SAR image.

[0064] The simulated SAR image generator 520 has a function of performing pseudo observation using the reconstructed three-dimensional information (data having information on reflection intensity and phase information at three-dimensional positions in a steady state) and the imaging conditions of one or more SAR images to be analyzed. Specifically, the simulated SAR image generator 520 generates a simulated SAR image, which is a complex image (a complex image showing a steady state) suitable for the imaging conditions of one or more SAR images to be analyzed. In other words, the pseudo observation means calculating an image predicted when imaging is assumed to be performed under the same imaging conditions as the above-described analysis SAR image.

[0065] The simulated SAR image generator 520 performs simulated observation for each of the imaging conditions of one or more SAR images to be analyzed. After performing the simulated observation one or more times, the simulated SAR image generator 520 outputs simulated SAR images corresponding to each of the one or more imaging conditions acquired.

[0066] The signal processing device 500 uses data containing information on reflection intensity and phase at three-dimensional positions in a steady state, i.e., three-dimensional information, and can therefore generate a simulated SAR image, which is a complex image showing the steady state according to the shooting conditions of the SAR image to be analyzed.

[0067] The three-dimensional information reconstruction unit 510 calculates three-dimensional information using a group of observation SAR images obtained by photographing an area using SAR.

[0068] In addition, the simulated SAR image generation unit 520 generates a simulated SAR image, which is a complex image showing a steady state suitable for the shooting conditions of the SAR image to be analyzed, using three-dimensional information in a steady state reconstructed using the group of observed SAR images and the shooting conditions of the SAR image to be analyzed.

[0069] Next, a description will be given of the operation of generating a simulated SAR image by the signal processing device 500. FIG.

[0070] The three-dimensional information reconstruction unit 510 executes three-dimensional information reconstruction processing (step S510), which is processing for reconstructing three-dimensional information of the region to be analyzed based on the observed SAR images.

[0071] Next, the simulated SAR image generating unit 520 of the signal processing device 500 executes a simulated SAR image generating process (step S520). The simulated SAR image generating process is a process for generating one or more simulated SAR images based on the imaging conditions of one or more SAR images to be analyzed and the reconstructed three-dimensional information. The simulated SAR image is an image in which simulated received signals observed under the same imaging conditions as the SAR image to be analyzed are recorded.

[0072] After executing the simulated SAR image generation process, the signal processing device 500 ends the signal processing.

[0073] Next, the three-dimensional information reconstruction process shown in Fig. 6 will be described with reference to Fig. 7. Fig. 7 is a flowchart showing the three-dimensional information reconstruction process executed by the three-dimensional information reconstruction unit 510.

[0074] First, the three-dimensional information reconstruction unit 510 calculates the steering vector r from the shooting conditions of each observation SAR image in the observation SAR image group. → (x → , n) is derived (step S511).

[0075] Next, the three-dimensional information reconstruction unit 510 repeatedly executes the process of calculating the complex reflection intensity distribution for each of the plurality of pixels.

[0076] Specifically, the 3D information reconstruction unit 510 selects one pixel from the observed SAR images for which a complex reflection intensity distribution has not yet been calculated. The selected pixel corresponds to a selected position in the observed SAR images.

[0077] Then, the three-dimensional information reconstruction unit 510 calculates the complex reflection intensity distribution α using the received signal of the selected pixel (selected position) in the observed SAR images and the steering vector for each of the observed SAR images. bg → (x → ) is calculated (step S512). → The complex reflection intensity distribution α of the pixel corresponding to bg → (x → ) is calculated.

[0078] The three-dimensional information reconstruction unit 510 repeatedly executes the process of step S512 until it has calculated the complex reflection intensity distribution for all pixels in the observed SAR images. That is, the three-dimensional information reconstruction unit 510 performs pixel loop processing. When the complex reflection intensity distribution for all pixels in the observed SAR images has been calculated, the pixel loop ends. When the pixel loop ends, the three-dimensional information of the target region has been reconstructed.

[0079] After exiting the pixel loop, the three-dimensional information reconstruction unit 510 outputs the calculated three-dimensional complex reflection intensity distribution as data having information on reflection intensity and phase at three-dimensional positions in a steady state, i.e., as three-dimensional information (step S514). The output three-dimensional information may be the three-dimensional complex reflection intensity distribution as described above, or may be three-dimensional point cloud data that is a set of points having information on reflection intensity and phase.

[0080] Next, the simulated SAR image generation process shown in Fig. 6 will be described with reference to Fig. 8. Fig. 8 is a flowchart showing the simulated SAR image generation process executed by the simulated SAR image generation unit 520.

[0081] First, the simulated SAR image generator 520 calculates the steering vector r from each of the imaging conditions of all input SAR images to be analyzed.→ (x → , n) is derived (step S521).

[0082] Next, the simulated SAR image generating unit 520 repeatedly executes the process of calculating a simulated complex signal for each of the imaging conditions of the SAR image to be analyzed, i.e., the simulated SAR image generating unit 520 executes imaging condition loop processing.

[0083] Specifically, in the imaging condition loop process, the simulated SAR image generating unit 520 selects one imaging condition that has not yet been used to generate a simulated SAR image from among the imaging conditions of the SAR image to be analyzed.

[0084] The simulated SAR image generation unit 520 repeatedly executes the process of step S522 until it has calculated simulated complex signals for all pixels in the observed SAR image group corresponding to the selected imaging conditions. That is, the simulated SAR image generation unit 520 performs pixel loop processing. When simulated complex signals for all pixels in the observed SAR image group have been calculated, the pixel loop ends. When the pixel loop ends, simulated complex signals for all pixels in the observed SAR image corresponding to the selected imaging conditions have been calculated. That is, a simulated SAR image corresponding to the selected imaging conditions has been generated.

[0085] Specifically, in the pixel loop, the simulated SAR image generating unit 520 selects one pixel for which a simulated complex signal has not yet been calculated from the pixels of the observed SAR images corresponding to the selected imaging conditions.

[0086] The simulated SAR image generator 520 generates the input complex reflection intensity distribution α bg → (x → ) and the steering vector r corresponding to the selected imaging condition → (x → , n) to find the position x corresponding to the selected pixel. → A simulated complex signal g sim (x → , n) (step S522). The simulated SAR image generating unit 520 calculates the simulated complex signal according to, for example, the following equation (3).

[0087]

[0088] The simulated SAR image generating unit 520 may calculate the simulated complex signal according to an equation other than equation (3).

[0089] When the pixel loop process has been executed for all the imaging conditions, the imaging condition loop is terminated. When the imaging condition loop is terminated, simulated SAR images of the target area corresponding to the imaging conditions of all the input SAR images to be analyzed have been generated.

[0090] After the imaging condition loop is completed, the simulated SAR image generator 520 outputs a simulated SAR image, which is a complex image (a complex image showing a steady state) suitable for the imaging conditions of the input SAR image to be analyzed (step S524). If there are multiple imaging conditions, multiple simulated SAR images will be output.

[0091] [Explanation of Effects of Reference Example] The three-dimensional information reconstruction unit 510 of the signal processing device 500 reconstructs data having information on reflection intensity and phase at each three-dimensional position in the steady state of the area to be analyzed, i.e., three-dimensional information, based on the group of observed SAR images stored in the SAR image storage unit 600. Furthermore, the simulated SAR image generation unit 520 generates one or more simulated SAR images based on the imaging conditions of one or more SAR images to be analyzed and the reconstructed three-dimensional information.

[0092] An advantage of using SAR tomography is that simulated complex signals indicating the reflection intensity and phase of each overlapping reflector are recorded in the simulated SAR image. Generally, even when using a digital elevation model (DEM), it is difficult to generate simulated complex signals accurately indicating the reflection intensity and phase of each overlapping reflector. However, the signal processing device 500 of the above-described reference example can generate a complex image indicating a steady state suitable for SAR image analysis. As a result, users who perform coherent change detection using the generated simulated SAR image can robustly detect changes even in layover areas. In other words, when detecting changes from the steady state of the SAR image being analyzed, the changes can be detected with high accuracy.

[0093] In the above-described reference example, the three-dimensional information reconstruction unit 510 uses SAR tomography as a means for calculating three-dimensional information. However, the three-dimensional information reconstruction unit 510 may use other means than SAR tomography as a means for calculating three-dimensional information that can reproduce reflection intensity and phase.

[0094] 9 is an explanatory diagram for explaining the change detection process when the SAR image showing the steady state includes information other than the steady state. As an example of the SAR image, the observed SAR image stored in the SAR image storage unit 600 is used. The observed SAR image group shown in FIG. 9 includes information A other than the steady state. １ Assume that there is an observed SAR image A containing:

[0095] In the above-described reference example, the three-dimensional information reconstruction unit 510 reconstructs three-dimensional information of the region to be analyzed based on the group of observed SAR images. １ Information A other than the steady state due to ２ As a result, the simulated SAR image generated by the simulated SAR image generating unit 520 is affected by the information A. ２ Information A other than the steady state due to ３ is reflected.

[0096] As an example of data used for change detection, coherence values ​​obtained from a simulated SAR image and an observed SAR image are taken. Figure 9 shows an example of a cross-correlation image, i.e., a coherence map, between the simulated SAR image and the observed SAR image. In the coherence map, the dotted areas have high coherence values. However, in area A, ４ The coherence value of area A is small. ４ The reason for the low coherence of is that information A ３ When a change detection process is performed using the coherence map shown in Fig. 9, there is a possibility that a false change detection occurs.

[0097] The signal processing device of the following embodiment can generate a complex image showing a steady state more suitable for SAR image analysis than the signal processing device of the above-mentioned reference example, and as a result, when detecting a change from the steady state of the SAR image to be analyzed, the change can be detected with higher accuracy.

[0098] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0099] Embodiment 1 [First Aspect of First Embodiment] Fig. 10 is a block diagram showing an example of the configuration of a signal processing device according to the first aspect of the first embodiment.

[0100] 10 includes a trusted 3D information reconstruction unit 110 and a simulated SAR image generation unit 120. A plurality of observed SAR images stored in a SAR image storage unit 600 are input to the trusted 3D information reconstruction unit 110. As in the reference example, an observed SAR image captured when there is a change from a steady state may be input to the trusted 3D information reconstruction unit 110.

[0101] When the trusted 3D information reconstructing unit 110 is included in a device other than the signal processing device 100, the signal processing device 100 includes only the simulated SAR image generating unit 120.

[0102] 4, the reliable three-dimensional information reconstruction unit 110 reconstructs and outputs three-dimensional information of the region to be analyzed. As in the reference example, the three-dimensional information may include information on temperature, displacement, etc. in addition to information on reflection intensity and phase.

[0103] In addition to the function of generating 3D information, the reliable 3D information reconstruction unit 110 has the function of calculating an index value indicating the reliability of the 3D information (hereinafter referred to as a reliability index value) and outputting the reliability index value.

[0104] An example of a method for calculating the reliability index value will be described.

[0105] The observed SAR image is an image in which the received signal is recorded. The reliable 3D information reconstruction unit 110 calculates a reliability index value by evaluating the discrepancy between the received signal and the received signal (predicted signal) predicted from the reconstructed 3D information for each of the reflection intensity and the phase. The reliable 3D information reconstruction unit 110 may calculate the reliability index value by evaluating the degree of likelihood of each estimated value in the reconstructed 3D information among possible values. Alternatively, the reliable 3D information reconstruction unit 110 may use both the evaluation of the discrepancy and the evaluation of the likelihood of the estimated value.

[0106] The reliable 3D information reconstruction unit 110 evaluates the discrepancy using, for example, the difference between the received signal and the predicted signal. When using the difference between the received signal and the predicted signal, squared difference or absolute difference, etc., is used. When evaluating the discrepancy, the reliable 3D information reconstruction unit 110 may evaluate a function that adds a term expressing the complexity of the reconstructed 3D information to the difference between the received signal and the predicted signal. Generally, 3D scatterers are often sparsely present, and a term generally referred to as a regularization term can be included as a term for determining whether such a solution is better. In particular, the L0, L1, or L2 norm can be used as the regularization term. When evaluating the discrepancy, the reliable 3D information reconstruction unit 110 may use cross-validation to evaluate the generalization performance of the reconstructed 3D information by using a received signal different from the received signal used for 3D information reconstruction as the received signal used to evaluate the difference from the predicted value.

[0107] The trusted 3D information reconstruction unit 110 outputs, for example, the parameters (variance, covariance matrix, confidence interval, etc.) of the posterior distribution of each estimated value obtained by Bayesian estimation or the posterior distribution as the likelihood of the estimated value. The trusted 3D information reconstruction unit 110 may calculate a reliability index value by evaluating a squared error or a loss function in optimal processing when reconstructing 3D information.

[0108] 4 , the simulated SAR image generator 120 has a function of generating a simulated SAR image based on the reconstructed 3D information input from the trusted 3D information reconstructor 110. The simulated SAR image generator 120 also receives input of imaging conditions for one or more SAR images to be analyzed. Similarly to the reference example, the simulated SAR image generator 120 generates a simulated SAR image, which is a complex image (a complex image showing a steady state) suitable for the imaging conditions for one or more SAR images to be analyzed.

[0109] In addition to the function of generating a simulated SAR image, the simulated SAR image generation unit 120 has the function of generating information (hereinafter referred to as reliability information) representing the reliability of each of the generated simulated SAR images using the reliability index value input from the reliable 3D information reconstruction unit 110.

[0110] An example of a method for calculating reliability information will be described.

[0111] The simulated SAR image generating unit 120 generates reliability information by statistical processing using the generated simulated SAR image and the reliability index value. For example, when the reliable 3D information reconstructing unit 110 uses Bayesian estimation, the simulated SAR image generating unit 120 evaluates the reliability of the simulated SAR image using the variance and covariance matrix of the posterior probability distribution of the obtained 3D information.

[0112] When evaluating reliability, the simulated SAR image generation unit 120 may use a predicted standard deviation that can be calculated for each imaging condition and each pixel of the simulated SAR image. In this case, the simulated SAR image generation unit 120 may use, for example, the predicted standard deviation itself as reliability information. Alternatively, the simulated SAR image generation unit 120 may compare the reliability information for each pixel with a predetermined threshold to identify pixels with low reliability, and use the identified result as reliability information.

[0113] Fig. 11 is an explanatory diagram for explaining an example of improved change detection processing when an observed SAR image showing a steady state includes information other than the steady state. Fig. 11 corresponds to an explanatory diagram for explaining processing by the signal processing device 100 and change detection processing using a simulated SAR image generated by the signal processing device 100. Specifically, part of the processing shown in Fig. 11 corresponds to processing executed by the trusted 3D information reconstruction unit 110 and the simulated SAR image generation unit 120.

[0114] As an example of the SAR image, the observed SAR image stored in the SAR image storage unit 600 is taken. The observed SAR image group shown in FIG. 11 includes information A other than the steady state. １ Assume that there is an observed SAR image A containing the following. The trusted 3D information reconstruction unit 110 reconstructs 3D information of the region to be analyzed based on the observed SAR images. The reconstructed 3D information is information A １ Information A other than the steady state due to ２ is affected by.

[0115] Furthermore, the reliable three-dimensional information reconstructing unit 110 calculates a reliability index value for the reconstructed three-dimensional information.

[0116] As described above, the reconstructed three-dimensional information is １ Information A other than the steady state due to ２ As a result, the simulated SAR image generated by the simulated SAR image generating unit 120 is affected by the information A. ２ Information A other than the steady state due to ３ is reflected.

[0117] In the coherence map shown in FIG. 11, the coherence value of the dotted area is large. ４ In this embodiment, the simulated SAR image generator 120 generates reliability information based on the reliability index, so that when the change detection process is performed, it is possible to avoid evaluating the coherence value for positions (pixels) with low reliability. As a result, when detecting a change from the steady state of the SAR image to be analyzed, the change can be detected with higher accuracy.

[0118] The motion detection process is performed by first motion detection section 130 or second motion detection section 140, which will be described later.

[0119] [Second Aspect of First Embodiment] FIG. 12 is a block diagram showing another example of the configuration of the signal processing device of the first embodiment.

[0120] 12 includes a trusted 3D information estimating unit 210 and a simulated SAR image generating unit 220. A plurality of observed SAR images stored in a SAR image storage unit 600 are input to the trusted 3D information estimating unit 210. As in the reference example, an observed SAR image captured when there is a change from a steady state may be input to the trusted 3D information estimating unit 210.

[0121] When the reliable 3D information estimation unit 210 is included in a device other than the signal processing device 200, the signal processing device 200 includes only the simulated SAR image generation unit 220.

[0122] The reliable three-dimensional information estimation unit 210 has the function of calculating a function that expresses information on how likely each value is as reflection intensity and phase at each point (each position) in a three-dimensional space having an azimuth direction, a range direction, and an elevation direction.

[0123] Hereinafter, this function may be referred to as “reliable 3D information.” Note that the information generated by the reliable 3D information reconstructor 110 shown in FIG. 10, i.e., the combination of the 3D information and the reliability index value, also corresponds to reliable 3D information.

[0124] The simulated SAR image generator 220 has the functions of the simulated SAR image generator 520 shown in Fig. 4. However, the above-mentioned reliability-assigned 3D information is input to the simulated SAR image generator 220 from the reliability-assigned 3D information estimator 210. The simulated SAR image generator 220 also receives input of the imaging conditions of one or more SAR images to be analyzed. Then, as in the reference example, the simulated SAR image generator 120 generates a simulated SAR image, which is a complex image (a complex image showing a steady state) suitable for the imaging conditions of one or more SAR images to be analyzed.

[0125] The simulated SAR image generation unit 220 has a function of generating a simulated SAR image and also a function of generating reliability information for each of the generated simulated SAR images. Note that the information (data) output by the simulated SAR image generation unit 220 is the same as the information output by the simulated SAR image generation unit 120. As described above, the information output by the simulated SAR image generation unit 120 is the simulated SAR image and reliability information.

[0126] Fig. 13 is an explanatory diagram illustrating another example of improved motion detection processing when an observed SAR image showing a steady state includes information other than that of the steady state. Fig. 13 corresponds to an explanatory diagram illustrating processing by the signal processing device 200 and motion detection processing using a simulated SAR image generated by the signal processing device 200. Specifically, part of the processing shown in Fig. 13 corresponds to processing executed by the trusted 3D information estimator 210 and the simulated SAR image generator 220.

[0127] As an example of the SAR image, the observed SAR image stored in the SAR image storage unit 600 is taken. The observed SAR image group shown in FIG. 13 includes information A other than the steady state. １ Assume that there is an observed SAR image A containing the following. The reliable three-dimensional information estimating unit 210 reconstructs three-dimensional information of the region to be analyzed based on the observed SAR images. Specifically, the reliable three-dimensional information estimating unit 210 generates reliable three-dimensional information (a function that expresses information on how likely each value is as reflection intensity and phase at each point). The three-dimensional information contained in the reconstructed reliable three-dimensional information is information A. １ Information A other than the steady state due to ２ is affected by.

[0128] [Method for Calculating Reliability-Assigned 3D Information in Second Aspect] The reliability-assigned 3D information estimation unit 210 calculates the reliability-assigned 3D information by using, for example, a posterior distribution. In this case, the reliability-assigned 3D information estimation unit 210 may, for example, use the posterior distribution itself as the reliability-assigned 3D information. The reliability-assigned 3D information estimation unit 210 may use parameters of the posterior distribution as the reliability-assigned 3D information. Examples of parameters of the posterior distribution include the mean, mode, variance, and confidence interval. Furthermore, the reliability-assigned 3D information estimation unit 210 may use a group of candidates for post-sampled 3D information (estimated values ​​of reflection intensity and estimated values ​​of phase) as the reliability-assigned 3D information.

[0129] The reconstructed three-dimensional information is information A １ Information A other than the steady state due to ２ As a result, the simulated SAR image generated by the simulated SAR image generating unit 220 is affected by the information A. ２ Information A other than the steady state due to ３ is reflected.

[0130] In the coherence map shown in FIG. 13, the coherence value of the dotted area is large. ４ The coherence value of is small. In this embodiment, the simulated SAR image generation unit 220 generates reliability information from the reliable 3D information estimation unit 210, so when the change detection process is executed, it is possible not to evaluate the coherence value for positions (pixels) with low reliability. As a result, when detecting a change from the steady state of the SAR image to be analyzed, the change can be detected with higher accuracy.

[0131] The motion detection process is performed by first motion detection section 130 or second motion detection section 140, which will be described later.

[0132] [Operation of the First Mode of the First Embodiment] Signal processing by the signal processing device 100 of this embodiment will be described below with reference to Fig. 14. Fig. 14 is a flowchart showing signal processing of the first mode executed by the signal processing device 100 shown in Fig. 10.

[0133] In the signal processing device 100, the reliable 3D information reconstruction unit 110 executes a 3D information reconstruction process (step S110). The 3D information reconstruction process is a process for reconstructing 3D information of a region to be analyzed based on a group of accumulated observation SAR images. In step S110, the reliable 3D information reconstruction unit 110 calculates a reliability index value of the reconstructed 3D information. The method for calculating the reliability index value has already been described.

[0134] Next, the simulated SAR image generator 120 executes a simulated SAR image generation process (step S120). The simulated SAR image generation process is a process for generating one or more simulated SAR images based on the imaging conditions of one or more SAR images to be analyzed and the reconstructed three-dimensional information. As described above, in the simulated SAR image generation process, the simulated SAR image generator 120 generates a simulated SAR image, which is an image in which simulated received signals observed under the same imaging conditions as the imaging conditions of each SAR image to be analyzed are recorded.

[0135] In step S120, the simulated SAR image generator 120 also performs a process of generating reliability information, which is calculated as described above.

[0136] Next, the three-dimensional information reconstruction process (step S110) shown in Fig. 14 will be described with reference to Fig. 15. Fig. 15 is a flowchart showing the three-dimensional information reconstruction process executed by the trusted three-dimensional information reconstruction unit 110.

[0137] First, the reliable 3D information reconstruction unit 110 calculates a steering vector r from the shooting conditions of each observation SAR image in the group of observation SAR images. → (x → , n) is derived (step S111).

[0138] Next, the reliable 3D information reconstructing unit 110 repeatedly executes the process of calculating the complex reflection intensity distribution and the process of calculating the reliability index value for each of the plurality of pixels.

[0139] Specifically, the trusted 3D information reconstruction unit 110 selects one pixel from the observed SAR images for which a complex reflection intensity distribution has not yet been calculated. The selected pixel corresponds to a selected position in the observed SAR images.

[0140] Then, the reliable 3D information reconstruction unit 110 calculates the complex reflection intensity distribution α using the received signals of the selected pixels (selected positions) in the observed SAR images and the steering vectors for each of the observed SAR images. bg → (x → ) is calculated (step S112). → The complex reflection intensity distribution α corresponding to the pixel bg → (x → ) is calculated.

[0141] The processes in steps S111 and S112 are the same as the processes in steps S511 and S512 in the reference example shown in FIG.

[0142] Furthermore, the reliable 3D information reconstructing unit 110 calculates the complex reflection intensity distribution α bg → (x → ) (step S113). That is, the trusted 3D information reconstructing unit 110 calculates a reliability index value of the reconstructed 3D information.

[0143] The trusted 3D information reconstruction unit 110 repeatedly executes the processes of steps S112 and S113 until it has calculated the complex reflection intensity distribution for all pixels in the observed SAR image group and the reliability index value for all pixels. That is, the trusted 3D information reconstruction unit 110 performs pixel loop processing. When the complex reflection intensity distribution and the reliability index value for all pixels in the observed SAR image group have been calculated, the unit exits the pixel loop. When the pixel loop is exited, the 3D information of the target region has been reconstructed and the reliability index value has been calculated.

[0144] After the pixel loop is completed, the reliable 3D information reconstruction unit 110 outputs the calculated 3D complex reflection intensity distribution as data containing information on reflection intensity and phase at 3D positions in a steady state, i.e., as 3D information, and also outputs the calculated reliability index value as 3D information (step S114). Note that the output reliability index value is a value obtained by combining the reliability index values ​​for each pixel calculated in the process of step S113.

[0145] Next, the simulated SAR image generation process (step S120) shown in Fig. 14 will be described with reference to Fig. 16. Fig. 16 is a flowchart showing the simulated SAR image generation process executed by the simulated SAR image generation unit 120.

[0146] First, the simulated SAR image generator 120 calculates the steering vector r from each of the imaging conditions of all input SAR images to be analyzed. → (x → , n) is derived (step S121). The process of step S121 is the same as the process of step S521 in the reference example shown in FIG.

[0147] Next, the simulated SAR image generating unit 120 repeatedly executes a process of calculating a simulated complex signal and a process of calculating the reliability of the simulated complex signal for each of the imaging conditions of the SAR image to be analyzed. That is, the simulated SAR image generating unit 120 performs imaging condition loop processing.

[0148] Specifically, in the imaging condition loop process, the simulated SAR image generating unit 120 selects one imaging condition that has not yet been used to generate a simulated SAR image from among the imaging conditions of the SAR image to be analyzed.

[0149] Then, the simulated SAR image generating unit 120 executes pixel loop processing. In the pixel loop processing, the simulated SAR image generating unit 120 selects one pixel for which a simulated complex signal has not yet been calculated from the pixels of the observed SAR images corresponding to the selected imaging conditions.

[0150] The simulated SAR image generation unit 120 repeatedly executes the processes of steps S122 and S123 until it has calculated the simulated complex signals and reliability information for all pixels in the observed SAR images corresponding to the selected imaging conditions. When the simulated complex signals and the reliabilities of the simulated complex signals for all pixels in the observed SAR images corresponding to the selected imaging conditions have been calculated, the pixel loop is terminated. When the pixel loop is terminated, the simulated complex signals and the reliabilities for all pixels in the observed SAR images corresponding to the selected imaging conditions have been calculated. That is, the simulated SAR images and reliability information representing the reliability of the simulated SAR images have been calculated.

[0151] In step S122, the simulated SAR image generator 120 calculates the input complex reflection intensity distribution α bg → (x → ) and the steering vector r corresponding to the selected imaging condition → (x → , n) to find the position x corresponding to the selected pixel. → A simulated complex signal g sim (x → , n). The simulated SAR image generator 120 calculates the simulated complex signal according to, for example, the above equation (3). The simulated SAR image generator 120 may calculate the simulated complex signal according to an equation other than equation (3).

[0152] The process of step S122 is the same as the process of step S522 in the reference example shown in FIG.

[0153] Furthermore, the simulated SAR image generator 120 calculates information indicating the reliability of the simulated SAR image (reliability information) by statistical processing using the reliability index indicating the reliability of the reconstructed three-dimensional information and the generated simulated SAR image (step S123). The method of calculating the reliability information is as already described.

[0154] When the pixel loop process is completed for all the imaging conditions, the simulated SAR image generator 120 exits the imaging condition loop. When the simulated SAR image generator 120 exits the imaging condition loop, simulated SAR images of the target region corresponding to the imaging conditions of all the input SAR images to be analyzed and reliability information for each simulated SAR image have been generated.

[0155] After the imaging condition loop is completed, the simulated SAR image generator 120 outputs a simulated SAR image, which is a complex image (a complex image showing a steady state) suitable for the imaging conditions of the input SAR image to be analyzed, and reliability information for each pixel (step S124). Note that if there are multiple imaging conditions, multiple simulated SAR images and reliability information are output.

[0156] [Explanation of Effects of the First Aspect of the First Embodiment] In this embodiment, the signal processing device 100 outputs reliability information indicating the reliability of the simulated SAR image in addition to the simulated SAR image. As a result, compared to the reference example, it is possible to prevent erroneous detection due to the inclusion of information other than that relating to the steady state, and the accuracy of change detection is improved.

[0157] [Operation of the Second Aspect of the First Embodiment] Signal processing by the signal processing device 200 according to another aspect of the present embodiment will now be described with reference to Fig. 17. Fig. 17 is a flowchart showing signal processing according to the second aspect executed by the signal processing device 200 shown in Fig. 12.

[0158] In the signal processing device 200, the reliable 3D information estimation unit 210 executes the 3D information reconstruction process using the calculation method already described (step S210). As described above, the 3D information reconstruction process is a process of calculating a function (reliable 3D information) that expresses information on how likely values ​​of reflection intensity and phase are at each point in a 3D space having an azimuth direction, a range direction, and an elevation direction.

[0159] The reliable three-dimensional information includes the three-dimensional information and information indicating the reliability of the three-dimensional information. That is, the reliable three-dimensional information essentially includes the three-dimensional information and information indicating the reliability of the three-dimensional information.

[0160] Next, the simulated SAR image generator 220 executes a simulated SAR image generation process (step S220). The simulated SAR image generation process is a process for generating one or more simulated SAR images based on the imaging conditions of one or more SAR images to be analyzed and the reconstructed three-dimensional information (substantially included in the reliable three-dimensional information). Specifically, the simulated SAR image generator 220 selects or generates a plausible simulated complex signal.

[0161] Next, the three-dimensional information reconstruction process (step S210) shown in Fig. 17 will be described with reference to Fig. 18. Fig. 18 is a flowchart showing the three-dimensional information reconstruction process executed by the trusted three-dimensional information estimation unit 210.

[0162] First, the reliable 3D information estimation unit 210 calculates the steering vector r from the shooting conditions of each observation SAR image in the group of observation SAR images. → (x → , n) is derived (step S211). The process of step S211 is the same as the process of step S511 in the reference example shown in FIG.

[0163] Next, the reliable 3D information estimation unit 210 repeatedly executes the process of calculating reliable 3D information for each of the plurality of pixels.

[0164] Specifically, the trusted 3D information estimation unit 210 selects one pixel in the observed SAR images for which the complex reflection intensity distribution has not yet been calculated. The selected pixel corresponds to a selected position in the observed SAR images.

[0165] Then, the reliable 3D information estimation unit 210 estimates reliable 3D information, i.e., a function that expresses information on how likely each value of reflection intensity and phase is based on the received signal of the selected pixel (selected position) in the observed SAR images (step S212).

[0166] The reliable 3D information estimating unit 210 repeatedly executes the process of step S212 until it has calculated the complex reflection intensity distribution for all pixels in the observed SAR image group. That is, the reliable 3D information estimating unit 210 performs pixel loop processing. When reliable 3D information for all pixels in the observed SAR image group has been calculated, the pixel loop ends. When the pixel loop ends, reliable 3D information for the target region has been generated.

[0167] After exiting the pixel loop, the reliable 3D information estimation unit 210 outputs reliable 3D information (step S214).

[0168] Next, the simulated SAR image generation process (step S220) shown in Fig. 17 will be described with reference to Fig. 19. Fig. 19 is a flowchart showing the simulated SAR image generation process executed by the simulated SAR image generation unit 220.

[0169] First, the simulated SAR image generator 220 calculates the steering vector r from each of the imaging conditions of all input SAR images to be analyzed. → (x → , n) is derived (step S221).

[0170] Next, the simulated SAR image generating unit 220 performs a shooting condition loop process, in which the simulated SAR image generating unit 220 selects one shooting condition that has not yet been used to generate a simulated SAR image from among the shooting conditions of the SAR image to be analyzed.

[0171] The simulated SAR image generating unit 220 then performs pixel loop processing. In the pixel loop processing, the simulated SAR image generating unit 220 selects one pixel for which a simulated complex signal has not yet been calculated from the pixels of the observed SAR images corresponding to the selected imaging conditions. The simulated SAR image generating unit 220 then performs estimation processing to estimate multiple simulated complex signal candidates estimated from the reliable 3D information and the likelihood of each simulated complex signal candidate, and select or generate a plausible simulated complex signal (step S222). When the estimation processing has been performed for all pixels in the observed SAR images corresponding to the selected imaging conditions, the pixel loop is terminated.

[0172] When the pixel loop process has been executed for all the imaging conditions, the imaging condition loop is terminated. When the imaging condition loop is terminated, simulated SAR images of the target region corresponding to the imaging conditions of all the input SAR images to be analyzed and reliability information for each simulated SAR image have been generated.

[0173] This is the same as in the first aspect of the first embodiment. In other words, the simulated SAR image generator 220 generates the same information (data) as the information (data) output by the simulated SAR image generator 120.

[0174] After the imaging condition loop is completed, the simulated SAR image generator 220 outputs a simulated SAR image, which is a complex image (a complex image showing a steady state) suitable for the imaging conditions of the input SAR image to be analyzed, and reliability information for each pixel (step S224). Note that if there are multiple imaging conditions, multiple simulated SAR images and reliability information are output.

[0175] [Explanation of Effects of the Second Aspect of the First Embodiment] In this embodiment, the signal processing device 200 outputs reliability information indicating the reliability of the simulated SAR image in addition to the simulated SAR image. As a result, compared to the reference example, it is possible to prevent erroneous detection due to the inclusion of information other than that relating to the steady state, and the accuracy of change detection is improved.

[0176] Second Embodiment Next, a second embodiment of the present invention will be described with reference to the drawings. Fig. 20 is a block diagram showing an example of the configuration of a signal processing device according to the second embodiment of the present invention.

[0177] 20 includes a trusted 3D information reconstruction unit 110, a simulated SAR image generation unit 120, and a first change detection unit 130. As shown in FIG. 20 , the signal processing device 101 receives an observed SAR image from a SAR image storage unit 600.

[0178] The functions of the trusted 3D information reconstruction unit 110 and the simulated SAR image generation unit 120 in this embodiment are the same as the functions of the trusted 3D information reconstruction unit 110 and the simulated SAR image generation unit 120 in the first embodiment.

[0179] The first change detection unit 130 receives one or more SAR images to be analyzed. The SAR images to be analyzed include received signals indicating reflection intensity information and phase information. The first change detection unit 130 also receives the imaging conditions for the one or more SAR images to be analyzed. The one or more imaging conditions are the same as the one or more imaging conditions input to the simulated SAR image generation unit 120. The first change detection unit 130 receives one or more simulated SAR images corresponding to each of the one or more imaging conditions from the simulated SAR image generation unit 120. Furthermore, the first change detection unit 130 also receives reliability information for each pixel from the simulated SAR image generation unit 120.

[0180] The first change detection unit 130 executes a change detection process. The change detection process is, for example, a correlation process using phase information of the simulated SAR image and phase information of the SAR image to be analyzed. When executing the correlation process, the first change detection unit 130 has a function of outputting a value calculated by the correlation process as a change detection result. The first change detection unit 130 may compare the value calculated by the correlation process with a predetermined threshold value and output information (data) indicating the presence or absence of a change based on the comparison result as the change detection result.

[0181] The correlation (correlation value) represents, for example, the degree of similarity between the SAR image to be analyzed and the simulated SAR image (hereinafter referred to as similarity). The similarity is represented, for example, by the distance between multiple images. The similarity may also be represented by an index other than the distance between multiple images. Hereinafter, the similarity may be referred to as correlation.

[0182] 21 is an explanatory diagram showing a specific example of the first change detection process executed by the first change detection unit 130. As described above, the reliable 3D information reconstructor 110 inputs the observed SAR images stored in the SAR image storage unit 600 and reconstructs 3D information. The reliable 3D information reconstructor 110 then outputs the reconstructed 3D information.

[0183] The imaging conditions of the SAR image to be analyzed and the reconstructed three-dimensional information are input to the simulated SAR image generator 120. The simulated SAR image generator 120 uses the input imaging conditions to generate a complex image showing a steady state that matches each imaging condition, i.e., a simulated SAR image.

[0184] The first change detection unit 130 detects whether or not a change from a steady state has occurred between the input SAR image to be analyzed and the simulated SAR image.

[0185] For example, the first motion detection unit 130 may detect the positions x → In this case, the coherence value γ(x → As described above, if there is no change from the steady state, the calculated coherence value will be large. On the other hand, if there is a change from the steady state, the calculated coherence value will be small.

[0186] The first change detection unit 130 detects, for example, each position x → The coherence value γ(x → ) is displayed two-dimensionally. If the SAR image being analyzed contains a location that has changed from the steady state, the location of the change is detected as a decrease in the coherence value.

[0187] Furthermore, as described above, when simulated SAR images are used, changes are robustly detected even in layover regions. Therefore, as shown in Figure 21, the first change detection unit 130 can identify change detection locations in layover regions. Note that in Figure 21, the dashed frame in the change detection results corresponds to the layover region.

[0188] As described above, the first change detection unit 130 of the signal processing device 101 detects changes that have occurred in areas of the SAR image to be analyzed by comparing the SAR image to be analyzed with the simulated SAR image.

[0189] For example, the first change detection unit 130 detects changes by calculating the degree of similarity between the SAR image to be analyzed and the simulated SAR image. The first change detection unit 130 may calculate the degree of similarity using phase information indicated by the SAR image to be analyzed and phase information indicated by the simulated SAR image. The degree of similarity may be, for example, a coherence value.

[0190] In this embodiment, the first change detection unit 130 also receives reliability information, but the input of the reliability information is omitted in Fig. 21. In other words, Fig. 21 shows processing that does not include processing based on reliability information, which will be described later.

[0191] Next, the operation of identifying a change detection point of the signal processing device 101 of this embodiment will be described with reference to Fig. 22. Fig. 22 is a flowchart showing the signal processing executed by the signal processing device 101.

[0192] In the signal processing device 101, the reliable 3D information reconstruction unit 110 executes the 3D information reconstruction process and the process of calculating the reliability index value of the reconstructed 3D information, as in the first embodiment (step S110).

[0193] In the signal processing device 101, the simulated SAR image generating unit 120 executes the simulated SAR image generating process and the process of calculating reliability information, similarly to the first embodiment (step S120).

[0194] Next, the first change detection unit 130 executes a first change detection process (step S130). The first change detection process is a process for detecting a change by executing a correlation process using phase information.

[0195] FIG. 23 is a flowchart showing the first change detection process.

[0196] In the first change detection process, the first change detection unit 130 repeatedly executes the process of step S132 (change detection process) until changes are detected for all pixels between the SAR image to be analyzed and a simulated SAR image generated under the same imaging conditions. That is, the first change detection unit 130 performs pixel loop processing. Below, we will take an example in which correlation processing is executed as the change detection process.

[0197] The first change detection unit 130 selects one pixel from the simulated SAR image for which a correlation has not yet been calculated. In step S132, the first change detection unit 130 calculates the correlation between the selected pixel in the simulated SAR image and a selected pixel in the SAR image to be analyzed. The selected pixel in the SAR image to be analyzed is a pixel at the same position as the selected pixel in the simulated SAR image.

[0198] Specifically, the first change detection unit 130 calculates the complex signal g of a selected pixel in the SAR image to be analyzed. obs (x → , n), and the simulated complex signal g of the selected pixel in the simulated SAR image generated under the same shooting conditions as the SAR image to be analyzed. sim (x → , n) The first change detection unit 130 calculates the correlation using the phase information.

[0199] For example, the first change detection section 130 uses a coherence value expressed by the following equation (4) as the correlation: In equation (4), E(·) represents an expected value.

[0200]

[0201] The first change detection unit 130 may use a value other than the coherence value that is calculated by correlation processing using phase information. For example, the first change detection unit 130 may extract phase information from the SAR image as a real value and perform correlation using the extracted real value. Furthermore, the first change detection unit 130 may calculate, as the similarity, the squared difference between the phase of a selected pixel in the SAR image to be analyzed and the phase of a selected pixel in the simulated SAR image.

[0202] In this embodiment, the first change detection unit 130 does not execute the process of step S132 for pixels in the simulated SAR image that have low reliability (step S131). The first change detection unit 130 determines whether the reliability is low based on reliability information input from the simulated SAR image generation unit 120. For example, the first change detection unit 130 determines that the reliability is low when the value indicated by the reliability information is smaller than a predetermined threshold. In other words, the first change detection unit 130 excludes pixels with low reliability indicated by the reliability information from the comparison target.

[0203] After the first change detection unit 130 executes steps S131 and S132 for the pixels of the simulated SAR image, it exits the pixel loop. After exiting the pixel loop, the first change detection unit 130 outputs a change detection result including the calculated correlation for each pixel (step S133).

[0204] When multiple simulated SAR images are input, first change detection unit 130 performs the first change detection process of step S130 for each of the input simulated SAR images.

[0205] Furthermore, the signal processing device 101 of this embodiment has a configuration in which a first motion detection unit 130 is added to the signal processing device 100 of the first aspect of the first embodiment (see FIG. 10 ). However, it is also possible to configure a signal processing device that performs motion detection processing similar to this embodiment by adding the first motion detection unit 130 to the signal processing device 200 of the second aspect of the first embodiment (see FIG. 12 ).

[0206] [Explanation of Effects] In this embodiment, in the signal processing device 101, the first change detection unit 130 calculates, using phase information, a correlation between a SAR image to be analyzed and a simulated SAR image generated under the same imaging conditions as the SAR image to be analyzed. By referring to the correlation calculated using the phase information, a user can easily detect a change from a steady state. Furthermore, by the first change detection unit 130 calculating the correlation between the simulated SAR image and the SAR image to be analyzed, a user can robustly detect changes even in layover regions.

[0207] Furthermore, in this embodiment, the first change detection unit 130 does not perform change detection processing on pixels with low reliability, which allows for more accurate detection of changes from the steady state even if the SAR image showing the steady state contains information other than that of the steady state.

[0208] As shown in FIG. 21, when first change detection section 130 projects the change detection result onto a two-dimensional map, the user can easily identify the location where the change was detected.

[0209] Furthermore, even if multiple changes occur in the target area over the observation period, the first change detection unit 130 can detect each change as a change from the steady state by calculating the correlation between the SAR images of the multiple analysis targets and the simulated SAR images corresponding to each change.

[0210] Furthermore, if multiple motion detection results are obtained, the first motion detection unit 130 can display the multiple motion detection results in chronological order, allowing the user to easily identify each motion detection location and each motion detection time.

[0211] Third Embodiment Next, a third embodiment of the present invention will be described with reference to the drawings. Fig. 24 is a block diagram showing an example of the configuration of a signal processing device according to the third embodiment of the present invention.

[0212] The signal processing device 102 shown in Fig. 24 includes a trusted 3D information reconstruction unit 110, a simulated SAR image generation unit 120, and a second change detection unit 140. As shown in Fig. 24, the signal processing device 102 inputs an observed SAR image from a SAR image storage unit 600.

[0213] The functions of the trusted 3D information reconstruction unit 110 and the simulated SAR image generation unit 120 in this embodiment are the same as the functions of the trusted 3D information reconstruction unit 110 and the simulated SAR image generation unit 120 in the first embodiment.

[0214] The second change detection unit 140 receives input of multiple SAR images to be analyzed. The SAR images to be analyzed include received signals indicating reflection intensity information and phase information. The second change detection unit 140 also receives input of the imaging conditions for the multiple SAR images to be analyzed. Each imaging condition is the same as each of the multiple imaging conditions input to the simulated SAR image generation unit 120. The second change detection unit 140 receives input of multiple simulated SAR images corresponding to each of the multiple imaging conditions from the simulated SAR image generation unit 120. Furthermore, the second change detection unit 140 also receives input of reliability information for each pixel from the simulated SAR image generation unit 120.

[0215] The second change detection unit 140 performs correlation processing using phase information for each pair of the SAR image to be analyzed and a simulated SAR image corresponding to the SAR image to be analyzed. Furthermore, the second change detection unit 140 outputs statistics obtained from the values ​​calculated for each pair as the change detection result. For example, the average or median of the values ​​calculated by the correlation processing using the phase information can be used as the statistics.

[0216] The second change detection section 140 may compare the statistic with a predetermined threshold value, and output information (data) indicating the presence or absence of a change based on the comparison result as the change detection result.

[0217] Next, the operation of identifying a change detection point in the signal processing device 102 of this embodiment will be described with reference to Fig. 25. Fig. 25 is a flowchart showing signal processing executed by the signal processing device 102.

[0218] In the signal processing device 102, the reliable 3D information reconstruction unit 110 executes the 3D information reconstruction process and the process of calculating the reliability index value of the reconstructed 3D information, as in the first embodiment (step S110).

[0219] In the signal processing device 102, the simulated SAR image generating unit 120 executes the simulated SAR image generating process and the process of calculating reliability information, similarly to the first embodiment (step S120).

[0220] Next, the second change detection unit 140 executes a second change detection process (step S140). The second change detection process executes a correlation process using phase information for each pair of the SAR image to be analyzed and a simulated SAR image corresponding to the SAR image, and detects changes by utilizing statistics obtained from the calculated multiple correlations.

[0221] FIG. 26 is a flowchart showing the second change detection process.

[0222] As in the second embodiment, the second change detection unit 140 executes pixel loop processing including the processing of steps S131 and S132. In this embodiment, the second change detection unit 140 executes pixel loop processing for all pairs of the SAR image to be analyzed and the simulated SAR image corresponding to the SAR image to be analyzed. Therefore, in this embodiment, image correlation is calculated for all pairs of the SAR image to be analyzed and the simulated SAR image corresponding to the SAR image.

[0223] Note that, similar to the first motion detection unit 130 in the second embodiment, the second motion detection unit 140 does not execute the process of step S132 for pixels in the simulated SAR image that have low reliability.

[0224] Next, the second motion detection unit 140 calculates statistics of the values ​​calculated in the correlation process for each pair (step S141), and outputs a motion detection result including the calculated statistics (step S142).

[0225] The signal processing device 102 of this embodiment has a configuration in which a second change detection unit 140 is added to the signal processing device 100 of the first aspect of the first embodiment (see FIG. 10 ). However, by adding the second change detection unit 140 to the signal processing device 200 of the second aspect of the first embodiment (see FIG. 12 ), it is also possible to configure a signal processing device that performs statistically-based motion detection processing similar to this embodiment.

[0226] [Explanation of Effect] When statistics are used as in this embodiment, it becomes possible to evaluate not only the high or low value of a single correlation but also the difference of each correlation from a typical value (e.g., the average value), thereby improving the robustness of change detection.

[0227] Furthermore, in this embodiment, the second change detection unit 140 does not perform change detection processing on pixels with low reliability, which allows for more accurate detection of changes from the steady state even if the SAR image showing the steady state contains information other than that of the steady state.

[0228] For example, the signal processing device 101 of the second embodiment and the signal processing device 102 of the third embodiment can be used to detect changes that occur in urban areas without requiring a person to physically visit them. The signal processing devices 101 and 102 can also be used to quickly detect when and where a change occurs in an area that is regularly monitored.

[0229] The signal processing devices 101 and 102 can also be used to monitor military bases and cities in areas of security concern, because users of the signal processing devices 101 and 102 can quickly detect the appearance of aircraft or vehicles in the monitored area simply by comparing the SAR image to be analyzed with the simulated SAR image.

[0230] Furthermore, the 3D information estimated by the reliable 3D information reconstruction unit 110 using SAR tomography may be stored on a server or the like. When the 3D information is stored on a server or the like, the user of the signal processing device 100, 102 can generate a simulated SAR image including the region of the change detection target. In other words, the user himself / herself does not need to reconstruct the 3D information.

[0231] Furthermore, for example, when creating training data for object detection in SAR images using machine learning, if the output of the first change detection unit 130 is added to the machine learning, the positions of objects added to the training data are limited to the output area of ​​the first change detection unit 130. In other words, the range of annotation is narrowed, and the cost required for learning is reduced.

[0232] Each component in the above embodiment can be configured as a single piece of hardware, or as a single piece of software. Each component can also be configured as multiple pieces of hardware, or as multiple pieces of software. Furthermore, some of the components can be configured as hardware, and the other parts can be configured as software.

[0233] Each function (each process) in the above-described embodiments can be realized by a computer having a processor such as a CPU (Central Processing Unit), a memory, etc. For example, a program for implementing the method (process) in the above-described embodiments may be stored in a storage device (storage medium), and each function may be realized by having the CPU execute the program stored in the storage device.

[0234] 27 is a block diagram showing an example of a computer having a CPU. The computer is implemented in a signal processing device. The CPU 1000 executes processing in accordance with a program (signal processing program) stored in a storage device 1001, thereby realizing the functions of the reliable 3D information reconstruction unit 110, the reliable 3D information reconstruction unit 210, the simulated SAR image generation units 120 and 220, the first change detection unit 130, and the second change detection unit 140 in the above-described embodiment.

[0235] The storage device 1001 is, for example, a non-transitory computer readable medium. Non-transitory computer readable media include various types of tangible storage media. Specific examples of non-transitory computer readable media include magnetic recording media (for example, hard disks), magneto-optical recording media (for example, magneto-optical disks), CD-ROMs (Compact Disc-Read Only Memory), CD-Rs (Compact Disc-Recordable), CD-R / Ws (Compact Disc-ReWritable), and semiconductor memories (for example, mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), and flash ROMs).

[0236] The program may also be stored in various types of transitory computer-readable media, to which the program is supplied, for example, via a wired or wireless communication path, i.e., via an electrical signal, an optical signal, or an electromagnetic wave.

[0237] The memory 1002 is realized by, for example, a RAM (Random Access Memory), and is a storage means for temporarily storing data when the CPU 1000 executes processing. A configuration is also conceivable in which a program held in the storage device 1001 or a temporary computer-readable medium is transferred to the memory 1002, and the CPU 1000 executes processing based on the program in the memory 1002.

[0238] A DSP (Digital Signal Processor) may be implemented in the signal processing device instead of the CPU 1000. The CPU 1000 and the DSP may also be implemented in the signal processing device.

[0239] Next, an overview of the present invention will be described. Fig. 28 is a block diagram showing the main components of a signal processing device according to the present invention. The signal processing device 10 shown in Fig. 28 includes a reliable 3D information reconstruction means (reliability-assigned 3D information reconstruction unit) 11 (implemented by the reliable 3D information reconstruction unit 110 or the reliable 3D information estimation unit 210 in the embodiment) that generates reliable 3D information including 3D information composed of estimated values ​​of reflection intensity and phase at a 3D position in a steady state reconstructed using an observed SAR image and information indicating the reliability of the 3D information, and a simulated SAR image generation means (simulated SAR image generation unit) 12 (implemented by the simulated SAR image generation units 120 and 220 in the embodiment) that uses the 3D information and the imaging conditions of the SAR image to be analyzed to generate a simulated SAR image, which is a complex image indicating a steady state suitable for the imaging conditions of the SAR image to be analyzed, and calculates reliability information indicating the reliability of the simulated SAR image.

[0240] In the signal processing device 10, the reliable 3D information reconstruction means 11 (e.g., the reliable 3D information reconstruction unit 110) calculates information indicating the reliability of the 3D information by, for example, evaluating the discrepancy between the received signal and the predicted signal predicted from the generated 3D information for each of the reflection intensity and phase.

[0241] In the signal processing device 10, a reliable 3D information reconstruction means 11 (e.g., a reliable 3D information reconstruction unit 110) evaluates, for example, how likely each estimated value in the generated 3D information is among the possible values.

[0242] In the signal processing device 10, the simulated SAR image generating means 12 (e.g., the simulated SAR image generating unit 120) generates reliability information, for example, by statistical processing using the generated simulated SAR image and information indicating the reliability of the three-dimensional information.

[0243] In the signal processing device 10, the trusted 3D information reconstruction means 11 (e.g., the trusted 3D information estimation unit 210) calculates, for example, a function that expresses, as trusted 3D information, information on how likely each value of the reflection intensity and phase at each position in 3D space is.

[0244] In the signal processing device 10, the simulated SAR image generating means 12 (e.g., the simulated SAR image generating unit 220) estimates, for example, a plurality of simulated complex signal candidates estimated from reliable three-dimensional information and the likelihood of each of the simulated complex signal candidates, and selects or generates a plausible simulated complex signal as reliability information.

[0245] The signal processing device 10 includes a change detection means (implemented by the first change detection unit 130 or the second change detection unit 140 in the embodiment) that detects changes that have occurred in an area of ​​the SAR image to be analyzed by comparing the SAR image to be analyzed with a simulated SAR image, and the change detection means may be configured to exclude pixels with low reliability represented by the reliability information from the comparison target (see step S131 in Figures 23 and 26).

[0246] The detection means may also detect the change by calculating the degree of similarity between the SAR image to be analyzed and the simulated SAR image.

[0247] Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

[0248] 10, 100, 101, 102, 200 Signal processing device 11 Reliable 3D information reconstruction means 12 Simulated SAR image generation means 110 Reliable 3D information reconstruction unit 120, 220 Simulated SAR image generation unit 130 First change detection unit 140 Second change detection unit 210 Reliable 3D information estimation unit 600 SAR image storage unit 1000 CPU 1001 Storage device 1002 Memory

Claims

1. a reliable three-dimensional information reconstructing means for generating reliable three-dimensional information including three-dimensional information composed of estimated values ​​of reflection intensity and phase at a three-dimensional position in a steady state reconstructed using an observed SAR image and information indicating the reliability of the three-dimensional information; a simulated SAR image generating means for generating a simulated SAR image, which is a complex image showing a steady state suitable for the imaging conditions of the SAR image to be analyzed, by using the three-dimensional information and imaging conditions of the SAR image to be analyzed, and calculating reliability information representing the reliability of the simulated SAR image; A signal processing device comprising:

2. The reliable three-dimensional information reconstruction means calculates information indicating the reliability of the three-dimensional information by evaluating a discrepancy between a received signal and a predicted signal predicted from the generated three-dimensional information for each of reflection intensity and phase.

2. The signal processing device according to claim 1.

3. The reliable three-dimensional information reconstruction means evaluates how likely each estimated value in the generated three-dimensional information is among possible values.

2. The signal processing device according to claim 1.

4. The simulated SAR image generating means generates reliability information by statistical processing using the generated simulated SAR image and information indicating the reliability of the three-dimensional information.

4. The signal processing device according to claim 2.

5. The reliable three-dimensional information reconstructing means calculates a function expressing information on how likely each value is as reflection intensity and phase at each position in three-dimensional space as the reliable three-dimensional information.

2. The signal processing device according to claim 1.

6. The simulated SAR image generating means estimates a plurality of simulated complex signal candidates estimated from the reliability-added three-dimensional information and the likelihood of each of the simulated complex signal candidates, and selects or generates a plausible simulated complex signal as the reliability information.

6. The signal processing device according to claim 5.

7. a change detection means for detecting a change occurring in an area of ​​the SAR image of the analysis target by comparing the SAR image of the analysis target with the simulated SAR image; The change detection means excludes pixels with low reliability represented by the reliability information from comparison targets.

7. A signal processing device according to claim 1, 2, 3, 5 or 6.

8. The change detection means detects the change by calculating a degree of similarity between the SAR image to be analyzed and the simulated SAR image.

8. The signal processing device according to claim 7.

9. The computer generating reliable three-dimensional information including three-dimensional information composed of estimated values ​​of reflection intensity and phase at a three-dimensional position in a steady state reconstructed using the observed SAR image and information indicating the reliability of the three-dimensional information; Using the three-dimensional information and the imaging conditions of the SAR image to be analyzed, a simulated SAR image is generated, which is a complex image showing a steady state suitable for the imaging conditions of the SAR image to be analyzed, and reliability information indicating the reliability of the simulated SAR image is calculated. Signal processing methods.

10. A computer comprising: generating reliable three-dimensional information including three-dimensional information composed of estimated values ​​of reflection intensity and phase at a three-dimensional position in a steady state reconstructed using the observed SAR image and information indicating the reliability of the three-dimensional information; Using the three-dimensional information and the imaging conditions of the SAR image to be analyzed, a simulated SAR image is generated, which is a complex image showing a steady state suitable for the imaging conditions of the SAR image to be analyzed, and reliability information representing the reliability of the simulated SAR image is calculated. A signal processing program for