Patent Application
20090243917
The present application is based on, and claims priority from, French Application Number 08 00349, filed Jan. 23, 2008, the disclosure of which is hereby incorporated by reference herein in its entirety.
The invention relates to the detection of the altitude of the isotherm at zero degrees Celsius, or zero isotherm, in particular on board a carrier (for example, an airplane), in order notably to inform the pilot or any other person navigating on board the carrier thereof.
The altitude of the zero isotherm corresponds to the boundary between the liquid phase and the solid phase of the atmospheric water. When a frozen hydrometeor (that is, a particle of water, in this case frozen) passes through this boundary, it starts to progressively melt.
It is particularly important for a pilot to know where the zero isotherm is located. This is because the risks need to be understood differently depending on whether the liquid or frozen phase applies, because the risks that are run are not the same. Moreover, the processing operations performed on the data also depend on the phase.
It should be noted that, beyond the zero isotherm, the dielectric properties of the hydrometeor are modified, as is its reflectivity. It will be recalled that, for a given point, the reflectivity corresponds to the ratio of the energy reflected by this point to the total incident energy.
This rise in reflectivity is translated into the appearance of a bright strip in high-resolution radar images. However, radars on board carriers have a resolution that is too low for it always to be possible to view this bright strip.
One known solution for determining the zero isotherm is to use temperature sensors, on board the carrier (U.S. Pat. No. 6,741,203 and U.S. Pat. No. 6,201,494) associated with a temperature decrease model and a measurement of the altitude of the carrier. However, the use of a temperature indication gives imprecise results. For example, the measurements can be very sensitive to the degree of humidity of the atmosphere.
Moreover, this solution is applicable only in relatively low latitude regions for which the altitude of the carrier is situated under the tropopause. Indeed, a reversal of the trend of the temperature is observed beyond this boundary: the model used is therefore no longer valid.
Other methods involve using data transmitted by sources external to the carrier: for example, weather bulletins transmitted by ground or other stations (U.S. Pat. No. 6,424,288). However, the transmission of the information is not always possible to fleet airplanes.
The invention aims notably to provide a solution to these problems.
One aim of the invention is to propose a method enabling an accurate evaluation of the altitude of the isotherm at zero degrees Celsius.
Another aim of the invention is to enable this evaluation to be performed on board a carrier in flight.
Another aim of the invention is to propose a method enabling a robust evaluation of the altitude of the isotherm at zero degrees Celsius.
To this end, according to a first aspect, there is proposed a method of processing a radar image, in particular obtained from an airborne radar.
According to a general characteristic, the method comprises an automatic evaluation of the altitude of the isotherm at zero degrees Celsius, called zero isotherm, using a processing of the reflectivity information conveyed by pixels forming all or part of the radar image.
In other words, the reflectivity information of the radar image or of a part of the radar image is used to determine the altitude of the zero isotherm. The detection of the altitude of the zero isotherm therefore results from processing of information and not from a simple observation of a bright strip.
Thus, even if the bright strip indicative of the altitude of the zero isotherm is not visible, the altitude of the zero isotherm can be evaluated from the processing of the reflectivity information.
Moreover, the evaluation is done automatically, for example each time a new radar image is acquired.
According to one embodiment, the processing can comprise a generation of at least two intermediate values of the altitude of the zero isotherm, from variations in the reflectivity information within the radar image, the value of the altitude of the zero isotherm being a function of the intermediate values.
Each radar image has N vertical slices, and the processing of the information can be performed on each vertical slice, one after the other. More specifically, the processing of the information is performed continuously over a series of vertical slices, without distinction. It will be noted that the term “radar image” should be understood to mean the volume scanned by the beam from the radar antenna, on each acquisition, the N vertical slices forming this volume.
According to one embodiment, the processing of the information can comprise:
the value of the altitude of the zero isotherm being evaluated from the main intermediate values of the altitude of the zero isotherm.
Preferably, at least three main intermediate values of the altitude of the zero isotherm are determined, at least two main intermediate values out of all the main intermediate values are combined, the standard deviation of each combination is calculated, the combination with the lowest standard deviation is selected, and if the standard deviation of the selected combination is less than a given threshold, the altitude of the zero isotherm is determined by calculating an average of the main intermediate values of the selected combination.
Preferably, the method can also comprise, for each vertical slice of the radar image, a search for so-called areas of interest formed by pixels conveying reflectivity information, the value of which is greater than a given threshold, the processing of the information being performed on the areas of interest of the vertical slices.
For example, a first method can comprise, for each vertical slice:
For example, a second method can comprise, for each vertical slice:
For example, a third method can comprise, for each vertical slice:
For example, a fourth method can comprise, for each vertical slice:
For example, a fifth method can comprise, for each vertical slice:
The application of several methods for determining a certain number of intermediate values of the zero altitude, values from which the zero altitude itself is determined, renders the detection of the altitude of the zero isotherm relatively robust.
According to another aspect, there is proposed a system for processing an image obtained from a radar, capable of implementing a method as described hereinabove.
For example, the radar can be airborne.
Still other objects and advantages of the present invention will become readily apparent to those skilled in the art from the following detailed description, wherein the preferred embodiments of the invention are shown and described, simply by way of illustration of the best mode contemplated of carrying out the invention. As will be realized, the invention is capable of other and different embodiments, and its several details are capable of modifications in various obvious aspects, all without departing from the invention. Accordingly, the drawings and description thereof are to be regarded as illustrative in nature, and not as restrictive.
The present invention is illustrated by way of example, and not by limitation, in the figures of the accompanying drawings, wherein elements having the same reference numeral designations represent like elements throughout and wherein:
Reference should be made to
The radar RD emits a signal in the form of a beam FS, so as to determine the position and the quality of the hydrometeors targeted by the radar beam FS.
An exemplary vertical slice CPi (or radar view) is represented in
The value of the reflectivity (or reflectivity information) corresponds to the amplitude associated with the pixel. The higher the altitude, the greater the reflectivity value, and vice versa. In other words, the darkest areas of the radar view signal the areas that have a high reflectivity value. In reality, the radar views are coloured: the dark grey areas are red, the mid-grey areas are yellow-orange and the light grey areas are green.
Reference is now made to
In this example, the steps 1 to 8 described hereinbelow are applied to each vertical slice CPi of a radar image.
In a first step 1, the data transmitted by the radar RD on board the carrier AV concerned are read, these data then being directly converted into Cartesian coordinates.
So as to improve the process, areas of interest are then detected in a step 2. These correspond to parts of the radar image where it is a priori easier to evaluate the altitude of the zero isotherm. This step provides a way of shortening the calculation time, and obtaining a more accurate result.
The processing of the reflectivity information of the vertical slice concerned is then performed. This processing comprises in this example the application of five distinct methods, steps 3, 4, 5, 6 and 7, to each of the vertical slices.
For each vertical slice, at least one intermediate value of the altitude of the isotherm at zero degrees Celsius, hereinafter called zero isotherm, is obtained. In this example, the methods 1, 2, 4 and 5 deliver an intermediate value of the altitude of the zero isotherm, whereas the method 3 delivers two thereof, by being based on two distinct criteria, as explained in more detail hereinbelow.
The number of methods and the method used are in no way limiting. Those skilled in the art will be able to use the appropriate number of methods according to the accuracy sought and the desired calculation speed.
Each of these methods is capable of exploiting the reflectivity information of the vertical slice being processed so as to deduce therefrom an intermediate value of the altitude of the zero isotherm.
These intermediate values of the zero altitude are then stored, step 8.
Once the steps 1 to 8 have been completed for the N vertical slices, a main intermediate value is selected for each method, step 9, or even two main intermediate values for one of the methods. Then, from the main intermediate values of the altitude of the zero isotherm, the value of the altitude of the zero isotherm is finally evaluated, step 10.
The latter first comprises a step 21 for selecting areas of maxima ZM, that is, the pixels px such that their reflectivity value rf(px) belongs to the top 20% of the reflectivity dynamic range Dyr. The expression “reflectivity dynamic range” should be understood to mean the range of the reflectivity values taken by the pixels of the vertical slice being processed. This range is delimited by the minimum and maximum reflectivity values that are found within the vertical slice being processed.
An Example A of vertical slice CPi is represented. On this slice, the grey areas represent the areas of maxima ZM1, ZM2, ZM3 and ZM4 selected on the slice, according to the abovementioned criterion.
Then, the areas of maxima that have a size called size (ZM), less than a given threshold, are deleted, step 22. Here, the threshold is a hundredth of the total number of pixels of the vertical slice concerned.
The Example B illustrates the deletion of the area ZM4 whose size is less than the set threshold.
From the remaining areas of maxima ZMi, an area of maxima called bottom area of maxima, ZMbs, is selected, step 23. This area is the one with the lowest average altitude (that is, the average of the altitudes of the pixels that form the area concerned). In the Example C, the bottom area of maxima ZMbs is the area ZM3. This area ZMbs serves as a reference thereafter.
In parallel, again on completion of the deletion of the areas of maxima that have too small a size, step 22, the areas of maxima ZM for which the altitude of the bottom limit alt[inf(ZM)] is less than a given threshold, in this case 4 km, are excluded, step 24. In other words, the areas of maxima situated at too high an altitude are deleted. In effect, the altitude of the zero isotherm is located in a band of altitude close to the ground, between 0 and 4 km. The application of this criterion leads to the deletion of the area ZM2, example D.
Finally, the areas of maxima ZM with an average altitude altavg (that is, the average of the altitudes of the pixels forming the area concerned) that is greater than a certain threshold dependent on the altitude of the bottom area of maxima, ZMbs, and whose upper limit is greater than that of the bottom area of maxima ZMbs, are deleted, step 25. In this example, the areas of maxima with an average altitude greater than that of the bottom area of maxima ZMbs to which 1000 m has been added are deleted.
As illustrated in the example E, this in this case leads to the deletion of the area of maxima ZM1.
The remaining areas of maxima then form the areas of interest ZI. In this example, only the area ZM3 is an area of interest, as represented in the example F.
The areas of interest ZI correspond either to precipitation areas, or to a bright strip signalling the location of the zero isotherm.
The method illustrated in
Then, a weighted average of the altitudes found is calculated, step 32. The weighting is a function of the distance of the pixel relative to the radar. The latter makes it possible to mitigate the blurring effect due to the reduced resolution because of the distance to the radar.
On completion of the method 1, a first intermediate value of the altitude of the zero isotherm, altitude 1, is obtained.
The histogram in question is constructed during a step 41. An exemplary histogram is represented in
The histogram comprises two reflectivity peaks, referenced peak1 and peak2. Between these two peaks, there is a reflectivity minimum Min.
During a step 42 (
The local minimum Min best separating the two determined peaks peak1 and peak2 is located, step 43. The reflectivity value associated with this minimum Min corresponds to the segmentation threshold, Threshold.
The vertical slice CPi is then binarized relative to the segmentation threshold, step 44. In other words:
in which x and y are the Cartesian coordinates of the pixels within the vertical slice concerned.
The pixels belonging to the boundary between the two areas are located and stored, step 45.
The pixels situated above the lowest average altitude of the areas of maxima ZM rejected(on selecting the areas of interest, step 2), are deleted as are the pixels adjacent to the rejected areas of maxima, step 46.
Ranges of pixels, that is, a series of points that are regular and as horizontal as possible, are then determined, step 47. Each range must be long enough to be taken into account (for example, have a number of pixels greater than a certain threshold). Moreover, the ranges with an average altitude (average of the constituent pixels) dispersed by more than 1500 m are not retained. In this example, a range of pixels is defined by a series of consecutive pixels with an altitude dispersion less than 500 m.
Once the ranges of points have been determined, the average altitude of the retained pixels is calculated while applying a weighting as a function of the distance of the pixel relative to the radar, step 48.
The result is validated if the number of pixels is sufficient, step 49. On completion of the step 49, a second intermediate value of the altitude of the zero isotherm, referenced altitude 2, is obtained.
More specifically, during a first step 51, the pixels of the vertical slice concerned associated with a reflectivity value belonging to the lower half of the reflectivity dynamic range are deleted.
Then, the columns that have too few valid pixels (in this case fewer than 1/10 th of the possible number of pixels per column) are deleted, step 52.
The average vertical reflectivity profile of the vertical slice concerned is then determined, step 53. One example of this type of profile is represented in
At two very close altitude values, there are a maximum gradient Grd and a reflectivity peak PicR. It is the altitudes that correspond to these two events Grd and PicR that are to be determined using the method referenced method 3.
Referring again to
Finally, the altitude of the strongest gradient of the profile, just above the altitude of the reflectivity peak, is sought, step 55. The latter constitutes another intermediate value of the altitude of the zero isotherm, referenced altitude 3a.
Referring now to
More specifically, during a first step, step 61, only the pixels with a reflectivity value that belongs to the last third of the reflectivity dynamic range Dyr of the vertical slice concerned are selected, this being done for the pixels px, such that their altitude is greater than the altitude of the bottom area of maxima ZMbs selected during the step 2.
Then, for each column col and until the final column colmax has been processed, the altitude of the strongest gradient of the column is sought.
The strong reflectivity variation between the pixels px9 and px10 is evidence of the appearance of a reflectivity gradient Grad1.
Referring again to
Then, the most regular range or ranges of pixels are determined, in the same way as for the method 2, step 64.
The average altitude of the pixels of the range or ranges determined in the preceding step is then calculated, step 65, this calculation being weighted as a function of the distance of the pixels to the radar.
Finally, the result is validated if the number of pixels that have been taken into account is sufficient, step 66. The intermediate value of the altitude of the zero isotherm delivered is referenced altitude 4.
The last method, method 5, used in this embodiment, is represented in
The latter is based on the search for the places where the variance of the reflectivity values is maximum, to determine the area where the reflectivity values are the least uniform. In effect, the zero isotherm separates two areas with very different average reflectivity values. The uniformity is therefore very low at the point of separation, implying a strong variance.
More specifically, during a first step 71, the pixels that do not belong to the area where the zero isotherm is located are deleted. This area is in this case defined by the average altitude of the lowest rejected area of maxima ZMrejbs and the average altitude of the lowest selected area of maxima, all in the step 2.
Then, the pixels with a reflectivity value that does not belong to the top third of the reflectivity dynamic range Dyr of the vertical slice concerned are deleted, step 72.
The remaining data are then windowed, for example using an analysis window of size [1;5]. The variance of the windowed data is calculated, step 73. In other words, the vertical slice concerned is covered pixel by pixel. The variance of the pixels situated inside the window is calculated. The result is then associated with the central pixel of the window.
The strongest calculated variances are retained, step 74. In this example, the retained variance values are those belonging to the final third of the variance dynamic range.
A weighted average of the altitudes of the pixels associated with the retained variances is then calculated, the weighting here again being a function of the distance of the pixel concerned to the radar, step 75. Finally, the result is validated if the number of pixels taken into account for the calculation of the average is sufficient, step 76. The value of the intermediate altitude deriving from the method 5 is referenced altitude 5.
When the methods have been applied to all the vertical slices, and the results have been stored, a main intermediate value of the zero altitude is selected for each method, step 9 in
A first step 91 comprises a selection of an altitude stratum (in this case 1000 m wide) where the most intermediate values are located for a given method, and this is done for each method.
Then, for each result obtained altitude 1, 2, 3, 3a, 4 and 5, a main intermediate value of the altitude of the zero isotherm is determined, step 92.
This value here corresponds to the central altitude of the above-mentioned altitude stratum, provided that the number of intermediate values of the zero altitude is sufficient. In this example, the chosen threshold is equal to 25% of all the possible intermediate values. If this condition is satisfied, the result is validated, step 93.
Once the main value has been calculated for a given result, the next result is selected, step 94, and this continues until all the results have been processed.
The altitude of the zero isotherm is ultimately calculated from the abovementioned main intermediate values, step 10 of
For this, all the possible combinations are made of three results out of the five results that are the main intermediate values of the results: altitude 1, 2, 3a, 4 and 5, step 101. The altitude 3 result is for the time being put to one side.
The standard deviation σ of each combination is then calculated, step 102. Then, the combination that has the lowest standard deviation is chosen, step 103. If the chosen standard deviation σc is less than 500 m, the average of the main intermediate values of the combination corresponding to the chosen standard deviation is calculated. This average corresponds to the altitude of the zero isotherm Iso 0 sought. The method is finished.
However, if the standard deviation σc is greater than 500 m, then all the possible combinations are made of three results out of the five results that are the main intermediate values of the results: altitude 1, 2, 3, 4 and 5. The altitude 3a result is in this case placed to one side, step 105.
As previously, the standard deviation σ of each combination is then calculated, step 106. The combination that has the lowest standard deviation is then chosen, step 107. If the chosen standard deviation σc1 is less than 500 m, the average of the main intermediate values of the associated combination is calculated. This average corresponds to the altitude of the zero isotherm Iso 0 sought. Otherwise, it is assumed that the measurements obtained are not sufficiently precise to provide a result for the altitude of the zero isotherm.
Obviously, the values of all the thresholds mentioned hereinabove are given purely as an indication. The latter can be adjusted according to the flying latitude of the airplane, the current season or even the geographic area where the airplane is located.
Moreover, it is possible to take account of the time trend of the value of the altitude of the zero isotherm, to determine each new value.
As a variant, it is advantageous to take account of a priori information for the automatic evaluation of the altitude of the zero isotherm. This information (outside temperature, weather bulletin) can be sent by a ground base during the flight of the airplane or even downloaded before departure.
It will be readily seen by one of ordinary skill in the art that the present invention fulfils all of the objects set forth above. After reading the foregoing specification, one of ordinary skill in the art will be able to affect various changes, substitutions of equivalents and various aspects of the invention as broadly disclosed herein. It is therefore intended that the protection granted hereon be limited only by definition contained in the appended claims and equivalents thereof.
| Number | Date | Country | Kind |
|---|---|---|---|
| 08 00349 | Jan 2008 | FR | national |
