METHOD FOR GEOREFERENCING REMOTE SENSING DATA

Abstract

The invention relates to a method for georeferencing remote sensing data of a remote sensing platform. According to the invention, a remote sensing dataset is received by the received sensing platform, which maps a visual range of the ground surface, and a georeferencing of the remote sensing dataset is determined using a reference dataset with known georeferencing.

Claims

1. A method for georeferencing remote sensing data of a remote sensing platform, recording a remote sensing data set by the remote sensing platform which depicts a visual range of earth's surface, a reference data set with known georeferencing is determined, wherein the reference data set depicts a portion of the earth's surface which overlaps at least partially with a visual range, during a first adjustment step a remote sensing data set and the reference data set are compared to each other, and in a second adjustment step, a morphology operation is applied to the remote sensing data set or the reference data set based on the comparison, repeating the first adjustment step and the second adjustment step until it is determined in the first adjustment step by comparing as a termination condition that the remote sensing data set and the reference data set differ from each other by less than a predefined threshold, in the event of the termination condition occurring, the georeferencing of the reference data set is set as the georeferencing of the remote sensing data set after application of the at least one executed morphology operation.

2. The method according to claim 1, wherein a spatial resolution of the remote sensing data set and a spatial resolution of the reference data set are brought into correspondence for comparison in the first adjustment step.

3. The method according to claim 2, wherein the resolution of the one of the remote sensing data set and the reference data set having the higher resolution is reduced to the resolution of an other of the remote sensing data set and the reference data set.

4. The method according to claim 2, wherein the remote sensing data set and the reference data set are values given in pixels of at least one measurement parameter, wherein for comparison in the first adjustment step a difference data set is created from the remote sensing data set and the reference data set having a number of pixels equal to a number of pixels of a data set with reduced resolution, wherein for amount of all positions of pixels of the difference data set, a pixel with position i of the difference data set has as its value a difference between values of a pixel of the remote sensing data set with the same position i and a pixel of the reference data set with the same position i.

5. The method according to claim 1, wherein the predefined threshold is a threshold value which is compared to value Diff=√(Σ_.sub.i(I.sub.A,i−I.sub.Ref,i).sup.2), wherein I.sub.A,i is a value of the pixel at position i of the remote sensing data set and I.sub.Ref,i is a value of the pixel at position i of the reference data set.

6. The method according to claim 1, wherein a morphology transformation includes at least one translation, at least one rotation and/or at least one perspective distortion of a corresponding data set.

7. The method according to claim 1, wherein a calibration of the remote sensing data set and a calibration of the reference data set are adapted to each other prior to the first adjustment step.

8. The method according to claim 1, wherein the remote sensing platform is a satellite, an unmanned aerial vehicle or a drone.

9. The method according to claim 1, wherein a preliminary georeferencing of the remote sensing platform is estimated for determination of the reference data set.

10. The method according to claim 1, wherein the remote sensing data set and the reference data set are recorded in a same spectral range.

11. The method according to claim 1, wherein the remote sensing data set and/or the reference data set are recorded in a infrared range.

12. The method according to claim 1, wherein the reference data set is recorded at a time interval from the remote sensing data of at most 24 hours.

Description

DESCRIPTION OF DRAWINGS

[0055] In the drawings:

[0056] FIG. 1 shows a remote sensing data set and a reference data set with different resolution

[0057] FIG. 2 shows a remote sensing data set, a difference image and with adjusted resolution, and

[0058] FIG. 3 shows reference images between the remote sensing data and the reference data.

DETAILED DESCRIPTION OF INVENTION

[0059] FIG. 1 shows a remote sensing data set on the left partial image, which is shown here as an example matrix with a large number of pixels. Each pixel includes a measured value, which is represented here by a grey scale. The remote sensing data set shown depicts a visual range of the Earth's surface, with pixels plotted along the path of the remote sensing platform over the Earth's surface in the vertical direction and pixels across the path on the horizontal axis. The remote sensing data in the left partial image do not yet have georeferencing.

[0060] The right partial image of FIG. 1 shows a reference data set whose georeferencing is known. The data of the reference data set are also plotted as pixels in the direction along the path and across the path. Again, each pixel includes a value of the measurement parameter, which can advantageously be the same measurement parameter as in the left partial image. The spatial resolution of the reference data set in the right image is lower than that of the remote sensing data set in the left image, so that the pixels here represent a larger portion of the Earth's surface. The object of the method according to the invention is to infer the georeferencing of the remote sensing data set from the georeferencing of the reference data set.

[0061] The reference data set shown in the right partial image was selected so that the portion of the earth's surface depicted by the reference data set overlaps with the visual range of the remote sensing data set, at least partially. The dashed region shows the sum of the pixels in the image of the remote sensing data that overlap geographically with the dashed pixel of the reference platform.

[0062] FIG. 2 in the left partial image shows the remote sensing data with a resolution that has been adjusted to the resolution of the reference data set. In the case shown, the resolution of the reference data was reduced for this purpose. Such interpolation can be done, for example, by simple weighted averaging or other known methods. It is assumed that for the currently assumed georeferencing, all pixels of the remote sensing data set that have an areal ratio in a geographically overlapping pixel of the reference data set are included in the interpolation (shown as a dashed square in FIG. 1 as an example for a pixel of the reference platform).

[0063] FIG. 2 in the right partial image shows a difference image obtained by subtracting the values of the pixels of the remote sensing data set from the values of corresponding pixels of the reference data set after adapting the resolution, wherein corresponding pixels are those that have the same position. A value can be calculated from the right partial image, for example as Diff=√(Σ_.sub.i (I.sub.A,i−I.sub.Ref,i).sup.2), wherein I.sub.A,i is the value of the pixel at position i of the remote sensing data set and I.sub.Ref,i is the value of the pixel at position i of the reference data set. In the example shown, this value can be Diff=1.63, for example.

[0064] Morphology operations can now be systematically applied to the remote sensing data set and in turn the difference image can be calculated. The steps of comparing and applying morphology operations are repeated until a termination condition arises, which is that the remote sensing data set and the reference data set differ by less than a predefined threshold. For this purpose, for example, the value Diff as defined above can be compared to threshold value. Morphology operations, for example, can be translation, rotation and/or perspective distortion of the image. Instead of the described value Diff, a cross-correlation or another metric describing the difference image can also be used. This value can be fed back to an optimization unit so that the value can be optimized iteratively, for example until a value of Diff=0 results. To this end, for example, changing of the input parameters can be carried out.

[0065] In some cases, the difference can depend very strongly on the relative pixel position and can fluctuate in the sub-pixel range, as will be shown in the following one-dimensional example: An IR data set is to be georeferenced, which includes a cooling tower of a power plant, which is about 1 pixel in size (in the reference data set) and surrounded by water. This can be exacerbated by the fact that under certain circumstances there may be no certainty in the absolute calibration, so that in such cases the absolute temperature values that would result from the IR data cannot be assumed to be accurate. If there is now an error of half a pixel in the original georeferencing estimate (after scaling the pixel size to the reference data set), half of the warm tower is located in a pixel area whose other half contains water. Thus, the tower's heat signal is drastically reduced and the expected high temperature value of the tower is not found. If the grid now moves through corresponding morphology transformations, in this case a pure translation, the signature of the cooling tower appears increasingly stronger until only a single pixel encompasses the tower. This example is also intended to illustrate that at coarse resolution it is not possible to assume fixed features against which to georeference. In many cases, these only result from the adaption of the resolutions. For such situations, the iterative method according to the invention to arrive at a suitable georeferencing is advantageous.

[0066] Such an iterative improvement is shown in FIG. 3. Here the improvement was achieved by a piecemeal translation of the remote sensing data set. Targeted optimization can be carried out, for example, using the Nelder-Mead simplex method, the Broyden-Fletcher-Goldfarb-Shanno method, the Davidon-Fletcher-Powell algorithm or using trust region approaches.

[0067] In FIG. 3 the left partial image shows the difference image corresponding to the right partial image in FIG. 2 after some iterations. Here, the value Diff as defined above has been reduced to Diff=0.55. Through further iterations, the value Diff can ideally be optimized to Diff=0, which is shown in the right partial image of FIG. 3.