EP3260808A2 - Method for offset correction of a weapon system - Google Patents

Abstract

The invention is based on a method for correction of the position of a weapon system (14), in which a projectile is fired from a gun (18) of the weapon system (14) in a target direction (30) onto an object (10), a direction (38). an object of action of the projectile at the object (10) is detected, the direction difference (40) between the target direction (30) and the direction of effect (38) is detected as a deposit (A) and a target direction of a later shot on the object (10) using the tray (A) is corrected. In order to achieve a fast and reliable offset correction, it is proposed that the direction of effect (38) be detected by taking an effective image (36) representing the projectile effect (34, 54) on the object (10) and the direction of the direction of the effect (38 ) is determined from the active image (36).

Description

The invention relates to a method for correction of a weapon system, in which a projectile from a gun weapon of the weapon system is fired at a target on an object, a direction of an action point of the projectile is detected by the object, the direction difference between the target direction and the Wirkpunakichtung detected as a shelf and a targeting of a later shot on the object is corrected using the bin.

Shots from guns are subject to a degree of variance with regard to their accuracy. In this case, a small scattering within a single shot sequence with unchanged orientation of the tube can be distinguished from a larger spread over longer periods of time and different weapons copies. As a rule, machine guns have a very small short-term spread, which is generally tolerable. The coarse scattering, however, caused by long-term changes, such as a transport of the weapon, temperature effects, wind changes from day to day and the like. The coarse scattering can thus lead in a combat situation to a systematically wrong filing, in which the barrel weapon with high accuracy, although reproducibly hits a small scattering area, this scattering area of the impacts but located next to the target object.

In order to detect such a storage error, a positioning shot at the target object or generally in a desired target direction can be fired. Now the impact of the positioning projectile is detected and thereby starting from the pipe to the point of impact a direction of effect. The distance from the impact to the target object or the direction difference from the direction of the direction of impact to the target direction can now be determined as a file. A later shot can be corrected using the clipboard so that this later shot hits the target object.

It is an object of the present invention to provide a method of storage correction of a weapon system which operates reliably and can be performed quickly.

This object is achieved by a method of the aforementioned type, in which according to the invention the direction of effect is detected by taking an effective image which represents the projectile effect on the object, and the direction of the effect is determined from the effective image. With the determination of the direction of effect from the active image, the directional difference between the target direction and the direction of the direction of impact can be detected automatically and swiftly - assuming that the targeted direction was assumed to be known. The alignment of the machine gun can be automatically corrected depending on the filing, so that the automatic correction of the determination of the direction of the direction of action in the active pattern to the direction correction of the gun barrel can be carried out very quickly and reliably.

The weapon system may include a machine gun for firing ammunition in salvo fire, in which automated shots can be fired in succession. Subsequent shots are expediently in an automated ratio to the previous shot, for example in the launch time difference and / or the automated stack correction. Conveniently, a control computer is provided which automatically aligns a gun weapon of the weapon system, for example, based on instructions from an operator.

The destination direction is expediently known, for example from a destination process. To perform the aiming operation, the object may be captured in a target image, and an operator may mark the imaged object, for example with a crosshair. The marker determines the direction from the weapon system to the object and thus the target direction. The target direction need not be a geoid absolute direction, but may be a relative direction, for example, to a reference direction of the weapon system, such as a picture corner, a Rohrausrichtung or the like.

The point of action of the projectile is a point at which the projectile unfolds an effect visibly, for example as an explosion, whose explosive lightning is visible. Also possible is a Aufaufwirbelung by a detonation pressure wave, a Cloud of smoke or other visible effect. The point of action lies on the object and can therefore be located directly on the object, for example in the form of an impact on the object, or in a neighborhood of the object. If the projectile is fired, for example, with a tempition and ignites in the vicinity of the object, for example, to achieve a scattering effect, so is the point of action of the projectile on the object in this case.

The filing does not have to be the absolute filing and can subsequently still be changed or corrected, for example, by a lead in the case of an accelerating moving target object. Also, another parameter may be used to correct the deposit, such as the measurement of variable wind, vibration of the weapon, anticipated acceleration of the target, and the like.

One possible method for correction of the position of the weapon system, ie the correction of the orientation of the weapon system on the basis of the determined deposit, may comprise the steps of determining the target direction, determining the direction of the direction of the effect, determining the deposit and the deposit correction of the machine weapon. To determine the target direction, an optical system with a camera can record the object, expediently in a regular, in particular continuous sequence of images. These images can be displayed to an operator on a screen so that they can track the object as in a movie. The operator can select the object and direct a mark on it, such as a crosshair. In the case of a relative movement of the object relative to the camera orientation, the operator can trace the marking to the object, or the object is automatically tracked so that the marking is automatically tracked to the object. For this purpose, the object is detected as such in the object image and recognized in the subsequent object images, so that the marking is tracked from object image to object image with the moving object. From the position of the mark in an object image, the target direction can be determined when shooting down the bullet, for example relative to the camera orientation or the object image or even absolutely, if an absolute orientation of the camera or the object image in space is known.

To capture the projectile effect, an object image can be examined for image abnormalities using image processing methods, which can be assigned to a projectile effect. If an image conspicuity is found in a given position of a projectile effect, the location of this image conspicuousness in the object image can be determined. The object image is now an active image that shows the effect of the projectile. The direction of the effect can now be determined from the position of the projectile effect or the image conspicuousness in the active image and may possibly be the situation itself.

Now the directional difference, for example by subtraction, can be determined from the target direction and the direction of the direction of the effect and used as at least provisional filing. The alignment of the weapon is now corrected using the bin, and a subsequent shot is aimed at the object. If necessary, the tray can be further corrected as described above.

The positioning shot can be made with the same or different ammunition as the later and corrective shot control shot, for example, to make the bullet effect more visible and concise. Also, a positioning shot from a salvo consist of several individual shots, which are directed one behind the other on the object and their effect is detected in each case in an active image.

The active image and the target image can overlap one another congruently, for example if the orientation of a camera receiving the two images is not changed between the two recording times. If the target direction is detected in the target image and the direction of effect is in the active image, then both directions can then be directly compared with one another, for example deducted from one another, so that the directional difference and thus the offset results therefrom. However, it may also be that the later active image is shifted to the target image by a movement of the receiving camera, which has been tracked for example by the operator or by auto-tracking the object. A movement of a vehicle carrying the weapon system can lead to a movement of the camera and thus a shift of the active image to the target image.

In order to be able to detect the directional difference in this case, in an advantageous embodiment of the invention, a target image which images the object is taken, and the target direction is defined in the target image. The target image can now be correlated with the later captured image using pattern recognition in the two images. In this way, the two images can be superimposed, so that an imaged in both images in Cover is brought. From the correlation, a relative position of the two images to each other can be determined. The directional difference can be determined from the relative position of the two images and the position of the target in the target image and the direction of the effect in the active image.

It may take several seconds between the launch and the detection of the projectile effect, depending on the distance of the targeted object. In this time, it may be that a target mark is carried along with the object, for example in subsequent object images. Decisive for determining the target direction is expediently the target direction at the time of firing of the projectile.

In this respect, in a further advantageous embodiment of the invention, a target image which images the object at the time of shooting is taken, and the target direction is defined in the target image. The target image is expediently recorded at the time of shooting. In the case of a series of images, the image which has the smallest time interval from the time of shooting can be used as the target image. After firing the projectile, a sequence of object images is recorded, whereby a target marker in these images can be carried along with the movement of the object. To determine the filing, only the target direction from the target image is expediently used. The target image can be an image in which the target or object is marked at launch.

It may happen that an image conspicuity in an object image is not so well recognizable that it can be clearly classified as a projectile effect, and even if it can be classified as a projectile effect, the site of action or the direction of the impact can not be determined exactly. Such a case can occur when an explosion flash disappears behind a dust wall or another influence looks similar to a bullet effect.

In order nevertheless to arrive at an evaluable result, it is proposed that a sequence of object images depicting object images is taken, the images of the sequence are examined for the projectile effect and a plurality of object images in which the projectile effect is pictorially found, as effect images on a development of the pictorial projectile effect in the impact images are evaluated. For example, from an expansion of a cloud of smoke or dust on one Turning point and also an impact time to be closed. It would also be possible to deduce from a course of a fire of a projectile on the Geschossart and thus to distinguish a marker shot from a shot from another gun. For example, the development may be compared to a waveform typical of a bullet waveform, in particular a marker fire of a particular type of ammunition for a positioning shot. For example, when shooting with a tracer bullet, the tracer can be tracked in the multiple impact images, and an impact point can be clearly assigned to the projectile.

Depending on the type of marker ammunition used in a positioning shot, the firing weapon system can be easily recognized by the target object. This creates the danger of counter-fighting. Depending on the marker ammunition, it may therefore be advantageous if it is only fired if previously the evaluation of a previous positioning shot is insufficient. If, for example, the combat scene shows that a positioning shot with normal combat ammunition is only insufficiently recognizable with regard to its point of effect, another positioning shot can be made in a subsequent shot with special marker ammunition.

During a fire, the object may defend itself by generating spurious signals, for example in the form of flares. In order not to confuse an interference signal with the projectile effect in such a case and thus to misrepresent the direction of the effect, it is advantageous if interference signals in a sequence of images or object images are recognized as such. Such a detection can be done by their temporal occurrence, for example, their signal start, a temporal waveform and / or a radiation spectrum of the noise is evaluated. Thus, the signals can be compared with a corresponding parameter of an expected projectile effect or with a corresponding parameter of known interference elements. If it is detected, for example, that a temporal occurrence of an interference signal can not result from a projectile effect, since, for example, the projectile may not yet have reached the object, the corresponding image conspicuousness can be recognized as an interference signal. The same applies to a temporal or spatial signal course. If, for example, a flare moves as it is not to be expected from the positioning floor, the image conspicuousness can also be detected as an interference signal.

Especially in heavy battles, it can happen that a complex battle is depicted in an object image. In the object image, for example, a muzzle flash is recognizable or the impact or the explosion of a foreign projectile from another weapon or the like. In this respect, an object image can have a plurality of image conspicuities, which are then to be examined as to whether the sought-after floor effect hides behind them. Such an evaluation can be computationally intensive and thus time consuming. Especially with explosive flashes of foreign bullets can also be a confusion occur, so that a filing is detected incorrectly.

With regard to one or more of these problems, it is advantageous if an image evaluation in an object image is spatially limited to an expectation range in which the projectile effect is expected. For example, an expectation range may be limited to the local environment of the nominal target point, for example around the target direction. In this respect, it is expedient to select the expectation range as a partial region in an object image which images the object and to weight an evaluation result differently within the expectation range, in particular to weight it higher than an evaluation result outside the expectation range. The size of the expected range is expediently at most half the image area of the object image, in particular at a maximum of 1/10 of the image area of the object image. An evaluation result is, for example, a calculation result as to whether an image conspicuity could represent the projectile effect, for example a probability that an image conspicuity represents the projectile effect. A weighting could be made such that a picture image outside the expectation range is classified with a lesser probability as a floor effect than the same image detail within the expectation range.

Depending on the size of the expectation range and with only a small probability that a projectile effect can be expected outside of the expected range, it may also be useful to restrict the area of image processing to the expected range, ie only to investigate the expected range for the projectile effect. Image abnormalities, such as an explosive flash, which are detected outside this range, then do not lead to incorrect alignment of the inserted tube weapon. In this respect, it is advantageous if, in order to determine the direction of the direction of the puncture, an image conspicuousness, which is a Projectile effect, whose situation lies outside the expectation range, is discarded.

An expectation range can be divided into several areas. Thus, the expected range may comprise a core area with a highest weighting and a border area with a lower weighting, possibly even a third outdoor area with an even lower weighting, but which is greater than the weighting outside the expectation range. It is also possible to have a continuous weighting curve which decreases with increasing distance to an expectation point in which the projectile effect of the positioning shot is expected.

The size of the expectation range in the object image is expediently not rigid, but can be selected as a function of one or more parameters. For this purpose, the parameter of a type-specific storage spread of the gun can be used. Also, the individual filing spread of the gun represents a meaningful such parameter. Gun-independent wind force and / or a wind direction can be considered. Movement of a vehicle carrying the gun during firing and / or a turret temperature are also parameters that may be taken into account.

A further advantageous embodiment of the invention provides that a time window within which the projectile effect is expected is determined. An evaluation of a series of object images can then be limited to those object images that lie at least partially within the time window. Generally speaking, a sequence of object images depicting the object may be taken, and images of the sequence may be examined for the floor effect, with a time of action of a floor effect within the time window being weighted higher than outside the time window.

From the occurrence of an image conspicuousness in one or more object images can be concluded that the time of action, ie the time at which a suspected projectile effect occurs. If this time of action lies outside the time window, then the probability is greater that the corresponding image conspicuity is a different effect, for example a different artillery, than if the effective time is within the time window. The time of action may be a time at which the projectile unfolds its effect, in particular begins to unfold. For example, if a light signal takes one Explosion or the burning of a Markiergeschosses a longer while and is mapped, for example, in several object images, so is the effective time at the beginning of this effect. This is true even if the time of action is not visible in the pictures, for example because a serve is hidden. From a subsequent expansion of a smoke or dust cloud, which was caused by the Geschossexplosion can be concluded, for example, on the basis of the expansion rate on the time of impact or the explosion of the projectile. This time can then be determined as the effective time.

Depending on the size of the time window, the distance of the object from the weapon system or other parameters, the evaluation of an image conspicuity that is outside the time window can be omitted, if it is clear that this image conspicuity can not be attributed to a projectile effect due to time constraints. In this respect, it is advantageous if, in order to determine the direction of effect from which the deposit is determined, an image conspicuousness, which could represent a projectile effect whose time of action lies outside the time window, is discarded.

The location and / or size of the time window can be determined from a distance of the object from the machine gun and / or a temping of the projectile. Thus, for example, the time window is placed symmetrically around the known from the tempation explosion time of the projectile.

To record the projectile effect, several object images can be taken in succession and evaluated according to the projectile effect or corresponding image abnormalities. One way to save on computation time and to avoid erroneous evaluations is that the inclusion of an active image, in which then the pictorial representation of the projectile effect is expected, is matched to the time window. Thus, the exposure period or integration period of the recording of the active image can be determined as a function of the position and size of the time window. Expediently, the temporal position of the integration lies within the time window. In this way, the evaluation can be limited to a single image. In particular, the integration period is set by an expected effective time, for example symmetrically around the expected time of action.

If the impact point and / or the explosion point of the projectile is obscured, for example by smoke, dust or an object, the determination of the action time from only one effective image is prone to error. In order to counteract this disadvantage, it is advantageous if a development of the pictorial projectile effect is evaluated in a sequence of active images. In this way it can be concluded, for example, from an expansion of a cloud of smoke by recalculation to its time of origin and thus to the time of action. Another possibility is that the burnup of a marker bullet is not visible from the beginning, for example, because the burn-up has started too small. However, if the completion of the burnup or a significant burnup phase is detected and, for example, a total duration of the burnup or the course of the burnup is known, the time of action, ie the time at which the bullet effect began, can also be determined in this way. In this respect it is advantageous if, from the retrospective analysis of the development, a time of action is determined. Subsequently, it can be checked whether the time of action lies in the time window.

In particular, if the active images are recorded in the infrared spectral range, it is advantageous if the development of the pictorial projectile effect is investigated. The enlargement of a heat range, for example, caused by the projectile explosion, can be tracked, and it can be concluded on the date of action by recalculation. In addition, it may be that the integration duration of the recording of an infrared active image is greater than the time window. In general, and thus independent of the time window, the time of action can only be determined inaccurately from a single image. By recalculating an expanding thermal cloud, however, one can infer the time and / or location of its formation.

Depending on the scenery, it may be that the image scene in an object image has strong signal structures and contrasts. Although such effects may possibly be reduced by a judicious choice of the spectral range of a receiving camera used, they may significantly increase the potential for error in the detection of a projectile effect. Thus, there are many types of false targets that erroneously conclude a bullet effect, so that the filing is calculated incorrectly. Examples include other explosions, burning buildings, all kinds of fires, muzzle flashes of weapons in the target area, and more.

In order to reduce or even suppress such false targets, it is advantageous to edit the object image with a smart image correction. Such editing is conveniently done using a previously captured object image. Thus, for example, between a shot of the projectile and an effective time of the projectile at least one object image are taken from the object, the later recorded image is processed using this object image, for example by enhancing image differences. In particular, such a processing can consist of an image subtraction, so that, for example, the object image and the active image are subtracted from one another. The projectile effect can now be determined from a remaining image conspicuousness and from this the direction of the effect. Even strong signal structures, if they change sufficiently slowly, can be reduced or even suppressed in this way. A period in which an object image used for image processing is used is expediently determined by the fluctuation of the signal structure in a series of object images, that is to say a temporal change of the signal structure.

In order to detect the occurrence of a projectile effect in a larger time range, it is advantageous if several object images are examined in succession by means of image processing, in particular image subtraction of at least one preceding, for example the respective preceding object image, on an image conspicuousness that could represent a projectile effect.

In order not to emphasize very short structural fluctuations, ie very rapid changes of individual image structures, it is advantageous if a reference image is obtained from a plurality of object images and the active image is processed on the basis of the reference image, for example by subtracting the reference image from the active image. The plurality of object images may be a sequence of object images. Although structures with high or fast fluctuation can not be eliminated in this way, they are weighted less and thus partially averaged out.

The reference image may be formed from an average of the object images or an interpolation from the object images. If, for example, a clear development of a structure can be recognized from several object images, this structure can be extrapolated by interpolation to the time of the active image to be examined are thus supplied to the image processing, for example, a picture subtraction.

If a reference image is created based on a plurality of object images, then it is advantageous if the number of object images from which the reference image is formed is determined as a function of the image acquisition frequency and / or a fluctuation of the scene around the object. The fluctuation of the scene can be determined from one or more preceding object images by image processing.

In order to detect even small image conspicuousnesses or mostly hidden projectile effects, it is advantageous if the difference image is processed with image processing in the form of a signal amplification, for example by at least a factor of 2, in particular a factor of 5. In order to avoid overdriving and / or errors in this case, it is advantageous if this image processing is only performed when no sufficient image conspicuousness, which could represent a projectile effect, has previously been found.

The effect of a projectile usually unfolds in several spectral ranges. Thus, an explosive flash in the visible range and also an expansion of a cloud of smoke in the visible range can be detected. The propagation of a thermal cloud can be sensed in the infrared spectral range. An explosion flash can be distinguished from flames very well in the ultraviolet range. In this respect, it is advantageous if a plurality of cameras that are sensitive in different spectral ranges are present and each camera receives at least one active image. From the impact images, either individually or from a combination of the results from the effects, a projectile effect can be determined and from this a direction of effect. In this way, one effect direction can be determined from all the active images of the cameras. It is also possible to first determine projectile effects, to assign matching projectile effects in the impact images to a positioning projectile and to determine only one direction of effect.

The pictorial development of the projectile effect usually unfolds differently in different spectral ranges. Thus, an explosive flash in the visible and / or ultraviolet range is very short, while the propagation of a thermal cloud takes much longer. The spread of a cloud of smoke or dust lasts even longer. Therefore, it is beneficial if the development of pictorial Projectile effect is evaluated in the impact images in each spectral range. The development of the pictorial projectile effects can then be compared with a signal course expected for the positioning shot, for example for an explosive flash, a spreading thermal cloud or an expanding cloud of smoke.

With a variety of signals around the object to be hit, it may be useful to limit the spectral range that is used to investigate the projectile effect. As a result, false signals can be suppressed. In this respect, it is advantageous if at least one spectral range is limited to a spectral band that is less than the factor of 1.5 by a marking color of the projectile. Depending on the frequency or spectrum, the factor can also be chosen to be smaller, for example 1.2 or even 1.1. The factor 1.5 means that the frequency of the upper end of the tape is at most 1.5 times as high as the frequency of the lower end of the tape.

In a further advantageous embodiment of the invention, a positioning shot is set with marking ammunition, which differs in their beam characteristic during the impact or the explosion of ammunition to combat the object. Generally speaking, ammunition with a special luminous characteristic can be used, which differs from the subsequent stock-corrected projectiles. In order now to detect the projectile effect of the positioning ammunition, it is advantageous to compare a signal characteristic of the projectile action found in one or more active images with a stored signal characteristic of the projectile. In this way, for example, the special positioning bullet can be identified in a tumultuous battle scene with many striking bullets with low probability of error.

An error probability can be further reduced if the signal characteristic of the projectile effect is examined in terms of its temporal course of the intensity in at least two different spectral ranges.

Another way to reduce false positives is to shoot a salvo of several floors onto the object. For each projectile, a projectile effect can now be detected and an effective time of the projectile effect can be determined. Now, for example, the time sequence of Acting times are checked for plausibility, for example by the temporal sequence is compared with the time sequence of the firing times.

Furthermore, it is advantageous if a time window in which the projectile effect is expected is determined for each projectile, and the position of the action time points relative to the time windows is determined. It can be inferred from the position of the time of action whether the projectile effect or the detected image conspicuousness is actually the projectile effect of the marking bullet (s). Thus, for example, an assignment value can be formed in each case from the position of the action times to the time windows, and a total assignment value can be formed from a plurality of assignment values. The total allocation value may indicate a probability of assignment of the projectile effects to the projectiles of the salvo.

In the case of a systematic error, for example if the distance to the object was determined incorrectly, it may be that the time of action is not all within an expected time window. In order to detect such a systematic error, it is advantageous if the position of all the action times is examined for a systematic positional shift to the time windows. Such a positional shift can then be corrected so that it has only a lower relevance in the calculation than without correction.

In order to be able to differentiate the projectile effects of a salvo of several projectiles from, for example, a machine-gun fire, it is advantageous if the time sequence of the firing times is coded irregularly. While control missiles are usually shot in a regular sequence when fighting an object, an irregular coding can be distinguished from this.

A further possibility for recognizing discrimination errors can be achieved by firing different marking bullets in the salvo, which produce a different bullet effect. All marking bullets used are expediently special projectiles for a Markierschuss, so that they differ so far from a later fired combat ammunition.

In order always to achieve a reliable recognition result in different combat scenes, it is advantageous if different modules for determining the projectile effect are available. Each module expediently pursues another strategy or another method for reducing recognition errors. The different modules can be used individually or in combination, expediently as a function of a combat situation, existing projectiles and / or existing detection systems.

There is also the possibility that the different modules are used together to determine the projectile effect, but depending on a combat situation, existing projectiles and / or existing detection systems, a different weighting of the modules is determined for determining the direction of effect.

• Ein Modul zur Begrenzung der Bildauswertung auf einen Erwartungsbereich, insbesondere die Umgebung des nominellen Zielpunkts bzw. zur Gewichtung eines Auswerteergebnisses in Abhängigkeit von der Entfernung der Bildauffälligkeit zum nominellen Zielpunkt.
• Ein Modul zur Bestimmung eines Zeitfensters um die erwartete Geschosswirkung und zur Reduzierung bzw. Gewichtung der Auswertung in Abhängigkeit von dem Zeitfenster.
• Ein Modul zur Bildbearbeitung, insbesondere zur Bildsubtraktion.
• Ein Modul zur Kombination mehrerer Sensorsysteme in unterschiedlichen Spektralbereichen.
• Ein Modul zur Erkennung unterschiedlicher Geschosse bzw. Spezialmunitionen zur Markierung.
• Ein Modul zur Verifikation der Trefferposition durch Korrelation mehrerer Explosionszeitpunkte.

Conveniently, at least two modules of the following module group are present:

• A module for limiting the image evaluation to an expected range, in particular the environment of the nominal target point or for weighting an evaluation result as a function of the distance of the image conspicuity to the nominal target point.
• A module for determining a time window around the expected floor effect and for reducing or weighting the evaluation as a function of the time window.
• A module for image processing, especially for image subtraction.
• A module for combining several sensor systems in different spectral ranges.
• A module for detecting different projectiles or special ammunitions for marking.
• A module for verification of the hit position by correlation of several explosion times.

The invention also relates to a weapon system having a target device for inputting and detecting a target direction into which a projectile from a weapon of the weapon system is fired at an object in a target direction. The weapon system expediently also comprises an optical system with a camera and an evaluation unit for determining a direction of an action point of the projectile on the object. The evaluation unit is expediently to prepared to determine a filing in the form of a directional difference between the target direction and the direction of the direction. The aiming device is expediently adapted to correct a target direction of a later shot on the object using the tray.

In order to achieve a fast and reliable automatic offset correction, it is proposed that the evaluation unit is prepared in accordance with the invention for detecting the direction of the effect by recording an effective image representing the projectile effect on the object and determining the direction of the effect from the effective image.

The description of advantageous embodiments of the invention given so far contains numerous features that are summarized in several dependent claims in several groups. However, these features may conveniently be considered individually and grouped together into meaningful further combinations, in particular when reclaiming claims, so that a single feature of a dependent claim can be combined with a single, several or all features of another dependent claim. In addition, these features can be combined individually and in any suitable combination both with the method according to the invention and with the device according to the invention according to the independent claims. Thus, process features can also be formulated formally as properties of the corresponding device unit and functional device features also as corresponding process features.

The above-described characteristics, features, and advantages of this invention, as well as the manner in which they are achieved, will become clearer and more clearly understood in connection with the following description of the embodiments, which will be described in detail in conjunction with the drawings. The embodiments serve to illustrate the invention and do not limit the invention to the combination of features specified therein, not even with respect to functional features. In addition, suitable features of each embodiment may also be explicitly considered isolated, removed from one embodiment, incorporated into another embodiment to complement it, and / or combined with any of the claims.

FIG 1

eine Objektszene mit Fahrzeugen und ein Waffensystem, das zur Bekämpfung eines der Fahrzeuge eingesetzt wird,

FIG 2

ein Zielbild, das das Waffensystem von der Objektszene aufgenommen hat und in dem ein Fahrzeug markiert ist,

FIG 3

eine Reihe von Objektbildern, die nach dem Zielbild aufgenommen werden,

FIG 4

ein Wirkbild, das einen Einschlag eines vom Waffensystem zum Fahrzeug abgeschossenen Geschosses zeigt,

FIG 5

einen zeitlichen Verlauf der Aufnahme eines Zielbilds, mehrere Objektbilder und mehrere Wirkbilder und darin sichtbare Ereignisse,

FIG 6

ein Ablaufschema zur gezielten Aufnahme eines Wirkbilds während eines Zeitfensters, in dem eine Geschosswirkung erwartet wird,

FIG 7

ein Verfahren zur Bildbearbeitung zur deutlicheren Erkennung einer Geschosswirkung,

FIG 8

ein Diagramm, in dem Sichtbarkeiten einer Geschosswirkung in drei verschiedenen Spektralbereichen über die Zeit dargestellt sind und

FIG 9

einen Zeitstrahl, auf dem der Verschuss einer Salve aus mehreren Einzelschüssen und die Zeitfenster dargestellt sind, in denen Geschosswirkungen der einzelnen Einzelschüsse erwartet werden.

Show it:

FIG. 1

an object scene with vehicles and a weapon system used to combat one of the vehicles,

FIG. 2

a target image that has picked up the weapon system from the object scene and in which a vehicle is marked,

FIG. 3

a series of object images taken after the target image,

FIG. 4

an impact picture showing an impact of a projectile shot down from the weapon system to the vehicle,

FIG. 5

a temporal course of the recording of a target image, several object images and multiple impact images and events visible therein,

FIG. 6

a flowchart for the selective recording of an active image during a time window in which a projectile effect is expected,

FIG. 7

a method for image processing for clearer detection of a projectile effect,

FIG. 8

a diagram in which the visibility of a projectile effect in three different spectral ranges over time are shown and

FIG. 9

a timeline depicting the firing of a salvo of several single shots and the time windows in which projectile effects of the individual single shots are expected.

FIG. 1 shows an object scene 2 in a landscape. The object scene 2 includes a number of vehicles, of which in FIG. 1 The following are shown by way of example: a rocket carrier 4 for launching ground-to-air missiles 44, two tanks 6, 8, a truck 10 and a number of other wheeled vehicles 12. In the foreground, a weapon system 14 is shown schematically, which is also shown only schematically in FIG Wheeled vehicle 16 is mounted. The weapon system 14 comprises a gun barrel 18 for projectile missiles and a target unit 20 with a plurality of cameras 22, of which FIG. 1 for simplicity, only two cameras 22 are shown. Furthermore, the weapon system 14 comprises an evaluation unit 24 for the numerical detection of a tray and for automatic storage correction of the gun 18.

At the in FIG. 1 represented scene of one of the vehicles from the object scene, such as the truck 10, to be combated by the weapon system 14. For this purpose, the target unit 20 generates by means of one or more of the cameras 22 a target image 26, in which the object scene 2 and in particular the vehicle to be controlled are shown. An operator of the weapon system 14 marks in the target image 26 an object 10 that is to be fought, for example, the front truck. This marking can be done via a target mark 28 in the target image 26, such as a crosshair, the sake of clarity in FIG. 1 is shown in the object scene 2 on the front wheeled vehicle. The target marker 28 is placed by the operator over the object 10 to be controlled or to the point at which a shot from the gun 18 is to be placed.

If the target marker 28 is set, for example by a corresponding command of the operator, the position of the target marker 28 in the target image 26 is detected by the target unit 20. This results in a target direction 30 from the weapon system 14 or the barrel weapon 18 to the target, in this case the object to be combated 10. The target direction 30 is only a relative target direction 30, since its absolute direction does not have to be known. Optionally, the orientation of the target image 26 receiving camera 22 is detected, so that the position of the target image 26 in the object scene 2 and thus the position of the target mark 28 in the object scene 2 is known. This results in an absolute target direction 30 from the weapon system 14 or the barrel weapon 18 to the target, in this case the object 10 to be controlled.

This target direction can now travel, for example, by a movement of the object 10 through the object scene 2 or a movement of the weapon system 14 in space. The target direction 30 can be moved in this case, for example by the position of the target mark 28 in the target image. The target mark 28 in the target image 26 can be tracked by the operator and thus held on the object 10 to be controlled. Another possibility is the automatic tracking of the object 10, so that the target marker 28 and thus the target direction 30 automatically remains held on the object 10. For this purpose, a series of target images 26 are recorded over time, with the original target image 26 by means of image-processed methods are compared, so that the object 10 is detected in the subsequent target images 26. Based on the original position of the target mark 28 on the object 10, the position of the target mark 28 of the new position of the object 10 in the object scene 2 is now adjusted automatically.

At the moment of a launch of a projectile from the gun 18 to the object 10, which is suitably controlled by an operator, the target direction 30 is stored at the moment of the launch. This target direction is used below to calculate a clipboard. Although previous objectives 30 prior to launch can be stored and used as a basis for other calculations, they do not initially play a role in the calculation of the filing.

The target image 26 displayed on a display unit of the target unit 20 is in FIG. 2 shown enlarged. It shows the target mark 28 on the front vehicle or object 10 at the time of the launching of the projectile in target direction 30 from the gun 18. After firing, a series of object images 32 are recorded, which are shown schematically in FIG FIG. 3 are shown. These object images 32 show the object scene 2, which changes slightly according to the passage of time. The object images 32 or at least a part thereof are examined for a pictorial representation of the projectile effect of the projectile fired from the gun 18. If such a pictorial projectile effect 34 is found in an object image 32, this object image 32 is used as the active image 36 for calculating the deposit.

FIG. 4 shows the active image 36, in which the projectile effect 34 next to the object 10 is visible. The projectile effect 34 is recognized as such in the active image 36, and the position of the pictorial representation of the projectile effect 34 in the active image 36 is recognized by the evaluation unit 24.

In FIG. 1 schematically shows how the evaluation unit 24 from the position L of the projectile action 34 in the active image 36 determines a Wirkpunktrichtung 38, R, in which the projectile effect 34, starting from a reference point, such as a picture corner or a point on the weapon system 14, as the mouth of the Pipe weapon, lies. This is analogous to the determination of the target direction 30 by detecting the orientation of the receiving camera 22 in the object scene 2 and thus the orientation of the active image 36 and further in conjunction with the position L of the pictorial representation of the projectile effect 34 in the active image 36th

From the direction difference 40, .DELTA.R between the direction of puncture 38, R and the target direction 30, the evaluation unit 24 determines a tray A with which the projectile projectile missed the object 10 or the desired target point therein. The tray A may consist of the direction difference 40, ΔR. From the filing A, the evaluation unit 24 now further calculates a filing correction K of the gun 18 in the form of a change in direction, in which the gun 18 is then pivoted. Subsequent shots at the object 10 are now fired in the corrected direction so that the object 10 is struck.

In the presentation off FIG. 1 For easier understanding, contents from the target image 26 and the active image 36 are shown together. FIG. 1 shows the position of the object 10 at the time of shooting, ie as shown in the target image 26. FIG. 1 However, also shows the projectile effect 34, which occurs later and is not visible in the target image 26, but only in the active image 36. In the active image 36, however, the object 10 has already moved on a bit, so that the filing on the target direction 30 at the time of shooting and the direction of action 38 is supported at the point of impact of the projectile, two directions 30, 38 which are detected at different points in time. This is in FIG. 1 schematically represented accordingly.

In the simplest case, the receiving camera 22 rests between the shooting time or the shot of the target image 26 and the effective time or the shot of the active image 36. The two images 26, 36 can be simply superimposed, so that the target direction corresponding to the position of the target mark 28 in the target image 26 and the direction of action according to the position of the projectile effect 34 in the active image can be easily tapped next to each other. As a rule, however, the camera 22 will move during the projectile flying time, so that the object scene 2 travels through the images 26, 36. This is based on the presentation 2 and FIG. 4 shown. The camera 22 has been slightly pivoted to the left during the flight time of the projectile of about 1.5 seconds by the movement of the supporting wheeled vehicle 16. In order to be able to maintain the filing calculation despite this movement, at least the active image 36 is correlated with the target image 26 by image processing, so that the two object scenes 2 can be superimposed, for example by static landscape points. The image shift is recognized from this and taken into account in the calculation of the direction difference 40, ΔR.

In order to determine the direction of effect 38, the object images 32 are examined for image abnormalities that could show the projectile effect 34. In this case, for example, all object images 32 taken one after the other are examined successively for a pictorial representation of the projectile effect 34. As in FIG. 1 is hinted at, the object scene 2 hereby show a variety of events that are easily confused with a pictorial representation of the projectile effect 34. In FIG. 1 is exemplified as from the two tanks 6, 8 flashes a muzzle flash. In addition, the object to be attacked 10 carries a machine gun whose muzzle flash in the upper area of the object 10 also lights up brightly. Further, in addition to the wheeled vehicles 12 a grenade impact 42 can be seen, which can also be easily confused with a projectile effect 34, especially when the projectile effect 34 consists primarily of a smoke effect or a raising of dust and stones. Next starts from the rocket carrier 4, a ground-air missile 44, the hot engine gases also cause a picture conspicuousness with a very strong signature. Therefore, it is not readily apparent to the evaluation unit 24 whether the grenade impact 42, the muzzle fire of the tank 6 or the muzzle flash of the MG on the object 10 is an effect of the projectile from the gun 18. If there is a mix-up, tray A will be miscalculated and subsequent shots from barrel weapon 18 will be misplaced.

In order to reduce the risk of confusion and thus the risk of error, the evaluation unit determines an expected range 46, which is shown by way of example in FIG FIG. 4 is shown. The expected range 46 is determined as a function of the target direction 30 and, for example, is placed symmetrically around the target direction 30, as in FIG FIG. 4 is exemplified, of course, the target direction from the target image 26 is used. In this expectation range 46, the impact or the explosion of the projectile is expected.

It is now possible to examine only the pictorial events within the expected range 46 and to ignore the image abnormalities outside the expected range 46. By way of example, it is shown that the muzzle flash of the tank 6 and the exhaust gas radiation of the ground-to-air missile 44 and the grenade impact 42 in this case can no longer lead to disruption or confusion. The risk of error is thereby significantly reduced. The expectation region 46 can be laid through all the object images 32, so that only the corresponding expected region 46 is examined for the pictorial representation of the projectile effect 34.

However, it is also possible to examine the entire area of the object images 32 for the pictorial representation of the projectile effect 34 and to weight only the examination result in accordance with its position to the expected area 46. If an image conspicuity is outside the expectation range 46, it will be weighted less heavily than if it were within the expectation range 46. Further differentiation is possible by using instead of a single-stage expectation range 46, as in FIG. 4 is determined, a multi-level expectation range is determined, which may consist of a core region and one or more areas lying around the core region, for example, a central region and an outer region. Depending on the position of the pictorial representation of the projectile effect 34 in the areas is weighted from the inside out ever less. The weighting can also be changed continuously, for example, with a growing distance of the projectile effect 34 from the target direction 30 continuously decreasing. In this way, further away from the targeted object 10 lying image abnormalities are not suppressed, but always weighted weaker, so that a closer impact finds greater consideration.

When choosing the size of the expected range 46, the evaluation unit 24 expediently takes into account both a type-specific storage spread of the gun and an individual filing spread of the gun, in this case the barrel weapon 18. While the type-specific storage spread can be specified by the manufacturer, an individual filing spread of the gun either also specified by the manufacturer or obtained from previous experiments. Another useful parameter, which can be taken into account in the size and also the position of the expected range 46, is the wind strength and in particular also the wind direction. This can be measured by the weapon system 14 and incorporated into the determination of the expected range 46. In accordance with the wind direction, the expected region 46 can be displaced somewhat downwind, for example relative to a symmetry about the target direction 30, and the size of the expected region 46 can be made dependent on the wind force. It likewise makes sense to move the gun or the vehicle 16 carrying the gun into the calculation of the expected range 46 to be included. The stronger the movement, the larger the expectation range 46, since a movement, such as shaking, can increase the spread.

Analogous to a spatial window, such as the expectation region 46, a temporal window can be defined in which a projectile effect 34 is expected. This is an example based on the representation FIG. 5 explained.

FIG. 5 FIG. 14 shows the recording periods 48 of a plurality of images of object images 32 on a timeline. Between the recording periods 48, a read-out period is indicated by hatching, in which no recording takes place and the corresponding detector of the image-recording camera 22 is read out. Of course, it is also possible to reduce or even eliminate the readout periods by sequential readout or other suitable methods. At the in FIG. 5 The hatched readout periods are mainly used to simplify the recognition of the method.

At a firing time t _A , the projectile is fired from the barrel weapon 18. In the corresponding recording period 48, the target image 26 is recorded. Subsequently, a plurality of object images 32 are recorded, as indicated by the interruption of the time beam based on the two bars.

The point in time t _E indicates an expectation time at which a projectile effect 34, for example an impact of the projectile on or at the object 10 or an explosion of the projectile in the air, is expected. At the expected time t _E , the evaluation unit 24 sets a time window 50 in which events, analogous to the spatial expectation range 46, are taken into account exclusively or taken into greater consideration. Depending on the method, therefore, the object images 32 lying ahead of the time window 50 are not taken into account or are not even recorded or taken into account with a lower weighting with regard to their image conspicuities.

A weighting or an assignment value can generally, even in the case of a spatial expectation range, be a parameter which is used in a decision as to whether an image conspicuity should be classified as a projectile effect. It is expediently a parameter whose size is decisive for the result of whether an image conspicuity should be classified as a projectile effect. The parameter can for example, a probability that the image conspicuousness is a projectile effect.

Such a distinction or weighting is shown by way of example with reference to an image conspicuity 52 in FIG FIG. 5 shown. The image conspicuity 52 is, for example, the muzzle-fire of the machine-gun on the truck 10. This image-conspicuity 52 continues for a while, as indicated by the double-headed arrow in FIG FIG. 5 is indicated by the image conspicuity 52, for example 50 ms. It extends in the example shown in three object images 32 and their recording periods 48. The entire time range of the image conspicuity 52 is outside the time window 50, so that the image conspicuity 52 is not considered or only weakly weighted as if it had been within the time window 50. This is all the more true, since the significant time of the image conspicuity 52 is to be placed at the beginning of the image conspicuity, on which therefore a corresponding explosion has taken place. In this respect, it is sufficient to use the time of the beginning of the image conspicuity 52 as the effective time with regard to the consideration or weighting.

The projectile effect 34, on the representation of the sake of clarity in FIG. 5 is omitted, is within the time window 50. It is taken into account accordingly, so that the location of their image conspicuity in the active image 36 is used to calculate the tray A.

Instead of a flash, as it would be caused by an explosion of a projectile and as in FIG. 4 for projectile action 34, it may be that a projectile effect 54 does not express itself or not only as lightning, but as a whirling up of dust, as a cloud of smoke or as spin-on of stones, as exemplified by the grenade impact 42 in FIG. 4 is shown. An explosion flash can be hidden or insufficiently visible.

One such example is in FIG. 5 shown. It can be seen that a projectile effect 54 is visible in an active image 36 which was recorded after the time window 50. This is difficult to recognize at first and may not be recognized as such. In the subsequent effective image 36, the projectile effect 54 is already greater, and in the subsequent active image 36, which is still larger, as in FIG FIG. 5 is indicated. It is, for example, the expansion of a thermal cloud, which is caused by the explosion of a projectile of the gun 18, and which is received by an IR camera 22 of the weapon system 14.

From the spatial change of the projectile effect 54 or the associated image conspicuousness in the object images two things can be concluded. Namely, on the one hand, that it is a projectile effect 54, so that the corresponding object images are now regarded as impact images, and on the other hand, a temporal origin, an action time t _W , be concluded. This is in FIG. 5 is indicated, in which the extent is interpolated back to an origin, which is indicated by the wedge-shaped lines. It is noteworthy that in an object image 32, which is even within the time window 50, no projectile effect 54 is visible. Maybe this is still too small or was covered by another object. Nevertheless, it can be concluded from the later impact patterns or the spatial development of the projectile effect 54, the time of action t _W. Is this within the time window 50, as in FIG. 5 is shown, it can be assumed that it is a projectile action 54 of a projectile from the barrel weapon 18. If the time of action t _W outside the time window 50, so the event can be weighted less or ignored, so it is not assumed that a corresponding projectile effect 54, but the image abnormalities are attributed to another event that is uninteresting for filing. It is important here that the projectile effects 54 do not necessarily have to indicate the time of action t _W in the active images 36, but that this can be determined from a temporal and spatial development of the projectile effect 54 and, as in FIG FIG. 5 may even be within a readout period in which no image capture takes place.

FIG. 6 shows a further possibility for efficient use of the calculated time window 50. Shown are recording periods 48 of object images 32, this time without the readout periods omitted for the image recording. Furthermore, the uninteresting image conspicuity 52 and the slow-growing projectile effect 54 are shown. The taking and / or evaluation of object images 32 whose recording period is completely in front of the time window 50 has been dispensed with. In this respect, the image conspicuity 52 is not discovered. The image capture of object images 32 is started only at the beginning of the time window 50, so that the first object image covers the beginning of the time window 50, as in FIG FIG. 6 is shown. It can be seen that two object images completely cover the time window 50. In this, however, the projectile effect 54 is not detected.

At the in FIG. 6 illustrated embodiment, the evaluation of the object images 32 only by the time window 50 would not lead to a positive result. It therefore makes sense to append a connection time window 56 to the actual time window 50 and also to examine this connection time window 56 for image abnormalities which might show the projectile effect 54. With this method, the projectile effect 54 can be successfully recognized in the form of the heat cloud. The size of the connection time window 56 is expediently to be selected such that image conspicuities from a dust upset, a smoke development and / or a heat development are reliably detected if their origin lies within the actual time window 50, as in FIG FIG. 6 is shown.

Another way to reduce the error is based on the presentation FIG. 7 explained. FIG. 7 shows the recording of a series of object images 32. Each of these object images 32 is merged with the respective preceding object image, which in FIG. 7 32-1 is compared, for example by image subtraction. Such image subtraction is indicated by the mark A - B in FIG FIG. 7 indicated. Consistent image features are thereby eliminated and only image changes are left as in FIG. 7 by way of example on a projectile effect 34 or its pictorial representation is shown.

In the event that no or no sufficient image conspicuousness is detected in the comparison image, the comparison image or difference image 58 can be processed with image processing in the form of a signal amplification. The correspondingly processed image 60 now shows the projectile effect 34 more clearly, so that its position in the active image 60 thus processed can be determined. For example, the location is determined relative to a reference point in the example FIG. 7 the picture corner on the top left. Accordingly, there are two coordinates, as in FIG. 7 is represented by the two double arrows.

Another possibility is that a group of object images 32 is grouped into a reference image 62, for example by averaging the object images 32 of the group or an interpolation from the object images 32. The reference image 62 is then compared with the next following object image 32, for example by image subtraction , as in FIG. 7 is indicated by the marking A - B. The further procedure can be carried out as described above.

The number of object images 32 from which the reference image 62 is formed may be made dependent on the image fluctuation within the object images 32. Here, the image fluctuation of the entire object image or only the expectation region 46 may be considered. The image fluctuation results from the change of the image contents from one object image 32 to the next. With high image fluctuation, ie rapid changes of the image contents, fewer object images 32 are processed to the reference image 62 than with a smaller image fluctuation.

• langwelliger Infrarotbereich,
• kurzwelligerer Infrarotbereich,
• visueller Bereich, in dem Blitze erkannt werden,
• langwelligerer visueller Bereich, der besonders zur Detektion von Rauch- oder Staubwolken geeignete ist, und
• Bereich innerhalb des ultravioletten Strahlungsspektrums.

Another factor contributing to the reduction of errors is when the object scene 2 is recorded by a plurality of cameras 22 which are sensitive in different spectral ranges. In the example off FIG. 8 Three cameras 22 are used, one being sensitive in the infrared spectral range, the other in the visual range and the third in the ultraviolet range. Of course, it is also possible to choose other spectral ranges, the representation of FIG. 8 is chosen only as an example. Meaningful are several areas of the group

• long-wave infrared range,
• short-wave infrared range,
• visual area where lightning is detected
• long-wave visual range, which is particularly suitable for the detection of smoke or dust clouds, and
• Range within the ultraviolet radiation spectrum.

The corresponding image abnormalities 64, 66, 68, 70 show a different course. While an explosive flash in the ultraviolet spectral range is detected only for a relatively short period of time, in the visual range the explosive flash is detected in addition to a dust swirl which expands relatively slowly and thereby becomes more conspicuous, as in FIG. 8 indicated in the middle drawing area. Although the thermal cloud detected in the infrared spreads out more slowly than the flash of light is noticeable, its size growth is faster than the smoke and dust turbulence detected in the visible range. In this respect, different characteristic conspicuity or image progressions result for each spectral range, ie intensity and size profiles as a function of time.

Each of the cameras 22 receives a corresponding series of active images 36 in their spectral range. As described in the previous figures Image abnormalities are examined for their possible representation of a projectile effect 34. In the active images taken after the time of action t _W , the image conspicuities 64, 66, 68, 70 are detected and the development of the projectile effect in the active images 36 occurs in each spectral range with one for the positioning shot. expected waveform in each spectral range compared. For this purpose, typical signal profiles of several types of ammunition are stored in the evaluation unit 24 for each spectral range. The ammunition type of the projectile that was fired is deposited, so that the typical signal courses for this projectile are called up and with the measured signal progressions, for example correspondingly FIG. 8 , can be compared.

The signal profile of the projectile effect 34 is thus examined in its time course of the intensity in all three spectral ranges. The image conspicuities 64, 66, 68, 70 are only assigned to the projectile effect 34 if the signal curve in all tested spectral ranges corresponds to the stored signal characteristic within a predetermined deviation.

Again, to reduce the error probability, it is advantageous if one or more of the spectral regions are limited to a relatively narrow spectral band. For this purpose, the bandwidths of two spectral ranges, namely the visual and ultraviolet spectral range are limited to a factor of 1.5, so that the frequency of the upper band end is 1.5 times as large as the frequency of the lower band end. The width of the infrared spectral band is limited by a factor of 2, so that the frequency of the upper band end is twice as large as the frequency of the lower band end.

To delineate the signal profile against, for example, a muzzle flash, the projectile fired by the gun 18 may be a special signal projectile which burns with a marker fire of its own. This marker fire differs significantly in its spectrum as well as the time course of the intensity of usual muzzle flashes or grenade explosions. The signal profile of such a marking fire, which, for example, first amplified in the ultraviolet range and then over a longer period of time in the visual area, in particular with a characteristic intensity and / or spectral course over time, can be compared with the stored signal waveform, so that the projectile effect 34 of the positioning shot is clearly recognizable as such. By the evaluation of a Radiation course over time in one or more frequency bands, a projectile effect 34, in particular a special positioning bullet or marking bullet, are also distinguished from interfering signals.

If, for example, flaws are thrown in defense of the object 10, for example flares, these are clearly visible in the object images 32 as image conspicuities, but their radiation characteristics are different from those of a normal projectile or a special positioning projectile. By comparing the time course of the image conspicuities caused by the interfering elements with the corresponding parameter in the expected projectile effect, the projectile effect 34 can be clearly distinguished from such interfering signals.

Another possibility for reducing the susceptibility to failure is that a volley of several projectiles from the barrel weapon 18 is fired at the object 10. For each projectile, a projectile effect 34 is determined. In several test stages a susceptibility to confusion can now be reduced.

FIG. 9 shows the shooting of a salvo of six floors in their time course. At time t ₁ , the first shot of the salvo is done. At the times t ₂ , t ₃ ,..., The further shots of the salvo are fired. If the gun 18 is not or only slightly moved during the salvo, the corresponding projectile effects 34 should be visible at the same location of their effective image 36. If, therefore, six projectile effects 34, or more generally spoken: as many projectile effects as projectiles were shot in the salvo, found at one point in the object scene 2, this is a strong indication that they are actually projectile effects 34 of projectiles from the barrel weapon 18 acts.

In such an examination, there is still confusion in, for example, a muzzle flash of a machine gun, such as a machine gun or another salvo firearm. Even impacts from such a foreign weapon could lead to confusion. In order to reduce the risk of such confusion, the time sequence of the firing times is coded irregularly. Out FIG. 9 it can be seen that the first three shots of the salvo follow each other regularly, then a longer break occurs and the last three shots of the salvo follow each other again regularly, but with a smaller time interval to each other. To be able to use this coding, each bullet is used or for each projectile effect 34 determines its effective time. If the temporal coding of the times of action equals the coding of the kills, then this is a strong indication that the actually sought bullet effects 34 were found.

Another possibility is that a time window in which the projectile effect 34 is expected is determined for each projectile. This too is in FIG. 9 shown. Six time windows 50 are calculated, which lie in their temporal course according to the temporal coding. The time interval between a firing time t _i and the corresponding time window 50 can be determined from a distance between weapon system 14 and object 10 or a temping of the projectiles of the salvo, ie a preset period of time after firing, after which the projectile explodes.

To verify the affiliation of the image conspicuities to the projectile effects 34, it is now possible to determine from the position of the action times to the time windows in each case an assignment value for each action time. In FIG. 9 It is shown that three smaller image abnormalities occur before the first time window 50. Their active times are correspondingly before the first time window, so that the assignment value of these action times to a bullet of the salvo is low. Time after the first time window 50 is again a picture conspicuousness. The fifth image conspicuity is also behind a time window 50. Again, the assignment value to a floor is small, albeit slightly larger due to the temporal proximity to the respective time window 50. Accordingly, all image abnormalities are dealt with. Overall, a total allocation value is formed from the assignment values. Since the individual image conspicuities are not in the time windows 50, the assignment values are low and also the total allocation value is low. The total allocation value is too low in this embodiment that the image abnormalities are assigned to the salvo.

However, the position of all times of action is then examined for a systematic positional shift to the time windows 50. It is noticeable here that the fourth, fifth and sixth image conspicuities, as well as the eighth, ninth and tenth image conspicuities lie behind the same time interval behind a time window 50. It can thus be a systematic shift of the effective times to the time windows. Calculatorisch now all image abnormalities are around the systematic postponement shifted temporally forward. It is striking that six of the ten in FIG. 9 displayed image abnormalities come to lie symmetrically in the time windows 50. Your assignment value will be adjusted accordingly. As a result, the total attribution value is above a threshold, so these six abnormalities are assigned to the volleys of the salvo. The attribution values of the first three abnormalities and the seventh image abnormality are still low, so it is assumed that they do not belong to the salve.

In order to reduce the likelihood of confusion even further, different bullets are shot in the salvo, which cause a different projectile effect. This is also different from later lost combat ammunition. The difference may be in a spectral difference, in an intensity of the resulting image conspicuity or in a type of image conspicuity, such as a photo or smoke development. In FIG. 9 Such different projectile effects are indicated by differently designed explosion flashes. In the salvo, every third storey is stronger than the previous two. This is off FIG. 9 visible, noticeable. Now, the image abnormalities on the different floor effect in terms of spectrum, time course and / or Bildauffälligkeitsart be examined. It is noticeable here that the coding corresponds to the fourth to sixth and seventh to tenth image conspicuousness of the different projectile effects of the projectiles from the salvo. In addition, it is striking that the seventh image conspicuousness falls through an even greater bullet effect from the salvo series. As a result, the image conspicuities which have been verified in terms of their projectile effect can be increased in their assignment values and the remaining image conspicuities receive even lower association values.

Depending on the object scene 2, combat situation or current order, the requirements for detecting the projectile effect 34 in the object scene can be very different. In order nevertheless to always come to a reliable detection of a projectile effect 34, the weapon system 14 is equipped with a plurality of modules 72 for detecting the projectile effects 34. These modules 72 are in FIG. 1 indicated.

The different modules 72 for determining the projectile effect 34 can be used individually or in combination according to an operator input or an evaluation result of the evaluation unit 24. It is also decisive which Cameras 22 are present in the weapon system 14, which spectral filters are present, which bullets are available for positioning shots and which bullets are used and the like. For example, one of the modules 72 determines an expected range in the object images or an active image in which the projectile effect 34 is expected. Another module 72 determines a time window 50 in which the floor effect 34 is expected. Another module 72 evaluates different radiation spectra of the projectile effect. Another module 72 controls the use of multiple cameras 22 in different spectral ranges for comparison of spectrally different image conspicuities 64, 66, 68. Another module 72 controls the use of special marker ammunition for positioning shots, which is different from later lost ammunition. This module 72 also controls the corresponding examination of the image conspicuities 64, 66, 68 for their similarity to stored radiation profiles. Another module 72 controls the use of a Markersalve with several successive projectiles and evaluates the image abnormalities accordingly.

LIST OF REFERENCE NUMBERS

2

object scene

4

rocketship

6

tank

8th

tank

10

Trucks / object

12

wheeled vehicle

14

weapon system

16

wheeled vehicle

18

barreled weapon

20

destination

22

camera

24

evaluation

26

target image

28

target marker

30

target direction

32

object images

34

Levels effect

36

Functional diagram

38

Action point direction

40

directional difference

42

grenades impact

44

Ground-to-air missile

46

Expected range

48

Recording period

50

Time window

52

image conspicuity

54

Levels effect

56

Connection time window

58

difference image

60

edit image

62

reference image

64

image conspicuity

66

image conspicuity

68

image conspicuity

70

image conspicuity

72

module

L

location

R

Action point direction

.DELTA.R

directional difference

A

filing

K

correction

t _A

Launch time

t _E

Expectation time

t _W

More time

t _i

Launch dates

Claims (16)

Method for correction of a weapon system (14), in which a projectile from a gun (18) of the weapon system (14) is fired into an aiming direction (30) on an object (10), a direction (38) of an action point of the projectile on the object (10), the directional difference (40) between the aiming direction (30) and the direction of sense (38) is detected as a bin (A), and a target of a later shot is corrected to the object (10) using the bin (A) becomes,
characterized,
in that the direction of effect (38) is detected by taking an effective image (36) representing the projectile effect (34, 54) on the object (10), and determining the direction of effect (38) from the effective image (36). Method according to claim 1,
characterized,
in that a target image (26) depicting the object (10) is recorded and the target direction (30) is determined in the target image (26), the target image (26) correlates with the later captured real image (36) using pattern recognition and from this a relative position the two images (26, 36) to each other is determined and using the relative position, the direction difference (40) is determined. Method according to claim 1 or 2,
characterized,
that a sequence of the object (10) imaging object images (32) is received, the object images (32) to be examined of the sequence on the projectile impact (34) and a plurality of object images (32) in which the projectile impact (54) is found pictorially, as impact images (36) on a development of the pictorial projectile effect (54) in the impact images (36) are evaluated. Method according to one of the preceding claims,
characterized,
that an expected range (46), in which the projectile impact (34, 54) expected when a partial area in an object (10) imaging the object image (32) is selected, and an evaluation result within the expected range (46) is weighted higher than an evaluation result outside the expectation range (46). Method according to claim 4,
characterized,
that the size and / or position of the expectation range (46) in the object image (32) is selected as a function of at least one of the parameters from the following group of parameters: design-specific storage spread of the gun, individual storage spread of the gun, wind force, wind direction, movement of the gun carrying Vehicle (16) during launch, temperature of a gun part. Method according to one of the preceding claims,
characterized,
in that a time window (50) in which the projectile effect (34, 54) is expected is determined, a sequence of object images (32) depicting the object (10) is taken, images of the sequence are examined for the projectile effect (34, 54) and an effective time (t _W ) within the time window (50) is weighted higher than outside the time window (50). Method according to claim 6,
characterized,
that the timing of the integration of the recording of the active image (36) using the time window (50) is fixed. Method according to claim 6 or 7,
characterized,
that a development of the pictorial projectile effect (54) is evaluated in a sequence of impact images (36), from the consideration of the development a time of action (tw) is determined, at which the projectile effect (54) began, and it is checked whether the time of action ( t _W ) lies in the time window (50). Method according to one of the preceding claims,
characterized,
in that at least one object image (32) is received by the object (10) between a shot of the projectile and an action time (t _W ) of the projectile, the active image (36) is processed using the object image (32), in particular the object image (32) is subtracted from the active image (36), and the direction of effect (38) from the difference image (58) is determined. Method according to one of the preceding claims,
characterized,
in that a plurality of cameras (22) which sense in different spectral regions are present and each camera (22) receives at least one active image (36), and from each of the active images (36) the direction of effect (38) is determined. Method according to one of the preceding claims,
characterized,
that a signal characteristic of the found in one or more active images (36) Levels effect is compared with a stored signal characteristic of the projectile. Method according to one of the preceding claims,
characterized,
that a salvo of several floors is fired at the object (10), for each floor a projectile effect (34, 54) and a time of action (t _W ) of the projectile effect (34, 54) are determined and the time sequence of the action times (t _W ) is compared with the time sequence of the firing times (t _A ). Method according to claim 12,
characterized,
that the timing of the firing times (t _A) is irregular coded. Method according to claim 12 or 13,
characterized,
that in the salvo different Markiergeschosse are fired, which cause a different projectile effect (34, 54), and which differ from a later fired combat munitions. Method according to one of the preceding claims,
characterized,
that different modules (72) for determining the projectile effect (34, 54) are present, which can be used individually or in combination, and are determined as a function of a combat situation, existing projectiles and / or existing detection systems, which module (72) determines the Projectile effect (34, 54), wherein one or more modules (72) of the following group are present: expected range (46) of the projectile effect (34, 54) in the active image (36), time window (50) for the projectile effect (34, 54), radiation spectrum of the projectile effect (34, 54), image subtraction of an object image (32) without projectile effect (34, 54) from the active image (36), use of multiple cameras (22) of different spectral ranges, use of marker ammunition, which differs from later lost ammunition is the use of a Marksalve several consecutively fired projectiles and their evaluation on several projectile effects (34, 54). A weapon system (14) comprising a target unit (20) for inputting and detecting a targeting direction (30) into which a bullet from a gun (18) is shot at an object (10), an optical system having a camera (22) and a Evaluation unit (24) for determining a direction (38) of an action point of the projectile at the object (10) and a file (A) in the form of a directional difference (40) between the target direction (30) and the direction of action (38), wherein the target device ( 20) is prepared to correct a target direction (30) of a later shot on the object (10) using the tray (A),
characterized,
in that the evaluation unit (24) is prepared to detect the direction of effect (38) by recording an effective image (36) representing the projectile effect (34, 54) on the object and determining the direction of effect (38) from the effective image (36) ,

Images