JPH05215848A - Image distortion correcting method and device therefor - Google Patents

Abstract

PURPOSE:To enable high-accuracy geometric distortion correction to a radar image by expressing the position of a spot (GCP) allowing the position to be indicated clearly on an image on line-pixel coordinates, using altitude information so as to obtain an orbital error. CONSTITUTION:A stored (2) uncorrected image is subjected to coarse geometric distortion correction by known orbital data. The corrected image is displayed (4), and a GCP position on the image is read by line-pixel coordinates. On the other hand, an altitude input part 5 reads GCP altitude from a digital topographical map on the basis of GCP latitude-longitude and converts the GCP position onto the coordinates on the image. The GCP position on the image is then compared with the theoretically computed position from the map to obtain a GCP error. This error is measured on plural GCPs, and the orbital data used for coarse geometric distortion correction is corrected by the orbital error estimated by calculating the average value of errors so as to form new orbital data, and fine geometric distortion correction is performed using this new orbital data. This is repeated until the convergence of the orbital data to obtain a correct fine geometric distortion corrected image.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention relates to a distortion correction device for correcting the distortion of image data acquired by, for example, a synthetic aperture radar (SAR) mounted on an artificial satellite, and more particularly, to a geometric correction device. It is related to academic distortion correction.

[Conventional technology]

 Figure 2 is a block diagram showing a conventional distortion correction device that corrects the distortion of the image data acquired by the optical sensor mounted on the artificial satellite. In the figure, (1) is the uncorrected image input section, (2) Is a data storage unit (3) that temporarily stores the uncorrected image input from the uncorrected image input unit (1) and inputs the uncorrected image from the data storage unit (2) to correct the geometric distortion of the image. The calculation unit (4) is an image display unit that displays the image whose geometric distortion has been corrected by the calculation unit (3).

 FIG. 3 is a flow chart showing an algorithm for correcting the geometric distortion of an image by the conventional distortion correcting apparatus.

 The conventional image distortion correction device is configured as described above and, for example, an uncorrected image recorded on a magnetic tape is input from the uncorrected image input unit (1) and stored in the data storage unit (2). Next, the uncorrected image stored in the data storage unit (2) is corrected for the geometric distortion of the image, that is, the distortion of the image shape according to the algorithm in Fig. 3.

 First, the computing unit (3) performs so-called coarse geometric distortion correction (6) that corrects the geometric distortion of uncorrected image data using known trajectory data. However, since the known orbital data contains decision errors, the image with coarse geometric distortion correction does not completely overlap the map exactly. For this reason, it is possible to use a point (Ground Control Point: GCP) that has a clear position on the map such as a bridge or an intersection and can clearly indicate the position on the image that has been subjected to the coarse-geometry correction, with even higher accuracy. to correct. First, the image with coarse geometric distortion correction is displayed on the image display unit (4), and the position of GCP on the image is read by coordinates (line, pixel) on the image (8). The relational expression showing the geometric distortion correction amount obtained when geometric distortion correction is performed is converted into the position (latitude, longitude) on the map [9]. The error is calculated by comparing the position (latitude, longitude) of GCP and the original position (latitude, longitude) of GCP read from the ground map (10). By measuring such an error for multiple GCPs and statistically processing the values, the error contained in the orbit data used for coarse geometric distortion correction can be calculated [11]. Therefore, if the geometric distortion correction is performed again using the trajectory data that has corrected this error, a highly accurate geometric distortion correction [12] can be performed, and an image that completely overlaps the map can be obtained.

[Problems to be Solved by the Invention]

 The conventional image distortion correction device as described above can accurately correct the optical sensor image (14). This is because the optical sensor does not depend on the altitude of the object as shown in Fig. 4, and the image position is fixed.

However, the radar image (13) is formed on the ground where the distance between the radar and the object (slant range) is equal, as shown in Fig. 4, so the image position depends on the altitude of the object. Therefore, in order to correct the geometric distortion of a radar image with a conventional image distortion correction device, the GCP position (line, pixel) on the image should be converted directly to the GCP position (latitude, longitude) on the map. There was a problem that an error occurred depending on the altitude for each GCP.

 The present invention has been made to solve such problems, and an object thereof is to obtain a distortion correction device capable of correcting geometric distortion with high accuracy for radar images.

[Means for Solving the Problems]

 The distortion correction device according to the present invention has an altitude input device so that the altitude as well as the latitude and longitude can be used as the position data of each GCP, and the GCP position is displayed on the line and pixel coordinates by using the altitude information. It expresses and calculates the orbital error.

[Action]

 In the present invention, since the latitude, longitude, and altitude are used as the position data of each GCP, the orbit used for the coarse geometric distortion correction and the GCP position (latitude, longitude, altitude) on the map are used for each calculation. The theoretical position that GCP should see on the image can be calculated. From the difference between the GCP position on the actual image and the theoretical position where GCP should be visible, the bias data of the trajectory data used for coarse distortion correction can be calculated. You can find the error. Accurate precise geometrical distortion correction can be performed by repeating geometrical distortion correction again several times using the orbital data in which the biased orbital error is corrected.

〔Example〕

 FIG. 1 is a block diagram showing an embodiment of the present invention, and (1) to (4) are exactly the same as the above conventional device. (5) is an altitude input section for inputting the altitude of GCP from the digital topographic map stored on a floppy disk or the like.

 FIG. 5 is a flow chart showing an algorithm for correcting the geometric distortion of an image by the distortion correcting apparatus according to the embodiment of the present invention.

 In the image distortion correction device configured as described above, the uncorrected image is input from the uncorrected image input unit (1) and stored in the data storage unit (2) as in the conventional image distortion correction device. Next, the uncorrected image stored in the data storage unit (2) is subjected to the coarse geometric distortion correction using the known trajectory data according to the algorithm of FIG. 5 (6) After that, the image after the coarse geometric distortion correction is displayed as an image. It is displayed on the section (4) (7), and the position of GCP on the image is read by coordinates (line, pixel) on the image (8).

On the other hand, in the image distortion correction device according to the present invention, the altitude input section (5) reads the altitude of the corresponding point from the digital topographic map based on the latitude and longitude of the GCP (15), and the GCP original position (latitude, latitude, (Longitude, altitude) and orbit data used for coarse geometric distortion correction are used to convert the GCP position to coordinates (line, pixel) on the image using the following equation (9), where the symbol for the following equation is As shown in Fig. 4, X is the theoretical position where GCP that can be calculated from the latitude / longitude and the orbit data used for coarse geometric distortion correction appears, X0 is the original position of GCP, h is the altitude of GCP, and H is The altitude of the satellite.

Thus, the GCP error is obtained by comparing the position (line, pixel) on the image of GCP obtained from the image with the position (line, pixel) theoretically calculated from the position on the map [10]. By measuring such an error for multiple GCPs and averaging the measured values for each line and pixel, the error included in the trajectory data used for coarse geometric distortion correction can be calculated (11 ). That is, the average error value in the line direction (satellite traveling direction) becomes the orbit error in the satellite traveling direction, and the average error value in the pixel direction (direction orthogonal to the satellite traveling direction) is the orbit error in the direction orthogonal to the satellite traveling direction. Become. With this orbital error, the orbital data used for the coarse geometrical distortion correction is corrected to new trajectory data, and the geometrical geometric distortion correction is performed using this data [12].

 By the way, the orbital error is obtained in this way, and the image with the corrected geometric distortion is obtained. However, since this calculation was not performed until this time, the value of x was calculated based on the trajectory data used for the coarse geometric distortion correction, and therefore, it was completed completely at It is not always possible to obtain an accurate orbital error and an image with precise geometric distortion correction. Therefore, the above process is repeated again for the geometric-geometry-distortion-corrected image obtained in this way to obtain the next trajectory data and the geometric-geometry-distortion-corrected image. When the orbital error is calculated in this way, the error is corrected to the orbital data, and the orbital error is calculated again, the orbital data converges to the correct orbital data, and the converged orbital data is used to correct the geometric distortion. The result is an accurate geometrical distortion-corrected image for the radar image.

 In the above embodiment, the altitude input section (5) is adapted to input a digital topographic map stored in a floppy disk or the like. This is, for example, a keyboard for inputting altitude information at a designated point. It may be one.

〔The invention's effect〕

 As described above, the present invention has an effect that a highly accurate geometric distortion-corrected image of a radar image, which has been impossible in the past, can be obtained by the simple structure of adding an altitude input section.

[Brief description of drawings]

 FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a block diagram showing a conventional image distortion correction device, FIG. 3 is a flow chart showing an algorithm of the conventional image distortion correction device, and FIG. FIG. 5 is a flow chart showing the algorithm of the present invention, which is an explanatory view of image acquisition by the radar and the optical sensor. In the figure, (1) is an uncorrected image input unit, (2) is a data storage unit that temporarily stores the uncorrected image input from the uncorrected image input unit (1), and (3) is a data storage unit (2 ) Is an operation unit that inputs the uncorrected image to correct the geometric distortion of the image, (4) is an image display unit that displays the image that has been geometrically corrected by the operation unit (3), and (5) is a floppy disk. This is an altitude input unit that inputs the altitude of the GCP from the digital topographic map stored on a disk or the like. In addition, (6) is the coarse geometric distortion correction, (7) is the image display, (8) is the GCP position reading on the image display, (9) is the GCP position converted to the position on the map, and (10) is the GCP error. Calculation, (11) is orbital error estimation by statistical calculation of GCP error, (12) is precision geometric distortion correction, and (15) is processing of GCP altitude input from digital topographic maps. (13) is the radar image and (14) is the optical sensor image. The same reference numerals in each figure indicate the same or corresponding parts.

Claims (2)

[Claims] 1. In the distortion correction of a radar image, coarse distortion correction is performed using known orbit data, and the coordinates on the image of a characteristic point are extracted from the rough distortion corrected image, and the altitude of the characteristic point is extracted. And the trajectory data are used to theoretically calculate the position where the feature point appears on the image, and the error of the trajectory data is estimated by comparing the position actually extracted from the image with the theoretically calculated position. Correcting the estimated trajectory error using the trajectory data corrected for the trajectory data used for the coarse distortion correction, and calculating the error of the trajectory data using the feature points again from the distortion-corrected image. An image distortion correction method characterized in that it repeats until the orbital data converges, and as a result, an accurate distortion correction image is obtained. 2. A device for correcting distortion of a radar image,
An image distortion correction unit, an image display unit that displays the distortion-corrected image and extracts the coordinates on the image of a characteristic point from the displayed image, and an altitude that inputs the altitude of the point with a specific characteristic. An image distortion correction device having a data input section.