Method for determining an actual distribution of fertilizer grains

Abstract

The invention relates to a method for determining an actual distribution of fertilizer grains comprising the steps of: designing at least one collecting device for fertilizer grains; spreading the fertilizer grains over the at least one collecting device by means of a fertilizer spreader; recording an image of the at least one collecting device, on which fertilizer grains have been spread, using a camera; localizing the fertilizer grains in the image, and calculating an actual distribution of the fertilizer grains on the collecting device and / or along a plurality of collecting devices, wherein the method comprises at least one improvement step which improves the localization of the fertilizer grains in the image.

Description

[0001] Method for determining an actual distribution of fertilizer grains

[0002] The invention relates to a method for determining an actual distribution of fertilizer grains and a system for recording an image in such a method.

[0003] When spreading fertilizer granules, the spreading pattern of a fertilizer spreader depends on the flow and flight characteristics of the fertilizer granules. These depend, among other things, on the grain size, grain shape, true density, bulk density, grain strength, moisture content, friction coefficient, and surface texture of the granules. In principle, there are recommended settings for different fertilizer spreaders, especially different centrifugal discs, which can be retrieved from databases or read from so-called spreading tables, taking the respective fertilizer type into account. However, deviations from the expected spreading pattern can occur, for example, due to variations in fertilizer quality, changes in the inclination of the spreader, spreading unit, and / or centrifugal disc(s), wind, moisture content of the fertilizer, changes in quantity, and / or segregation of the grain size fractions. Consequently, the actual distribution of the fertilizer, particularly with regard to lateral distribution, must be verified in practical application.

[0004] For this purpose, it is known, for example from EP 2 923 546 B1, to use an adhesive mat plate to catch and hold spread fertilizer granules, to spread fertilizer granules over these and to determine the distribution of fertilizer granules that have arrived on the adhesive mat plate.

[0005] The invention is based on the object of providing an improved method for determining an actual distribution of fertilizer grains and a system for recording an image in such a method.

[0006] The invention comprises a method according to claim 1 and a system according to claim 16.

[0007] The method for determining an actual distribution of fertilizer granules comprises the steps of laying out at least one collecting device for fertilizer granules; spreading the fertilizer granules over the at least one collecting device by means of a fertilizer spreader, in particular a centrifugal fertilizer spreader; recording an image of the at least one sprinkled collecting device with a camera; locating the fertilizer granules in the image, and calculating an actual distribution of the fertilizer granules on the collecting device and / or along a plurality of collecting devices, wherein the method comprises at least one improvement step that improves the localization of the fertilizer granules in the image.

[0008] At least one collecting device is designed in the process. In particular, one, two, three, or more collecting devices can be designed. The at least one collecting device is typically designed to be located in an area where fertilizer granules are expected to impact during spreading.

[0009] A collecting device can, in particular, comprise or be an adhesive mat as described in EP 2 923 546 B1. Alternatively, a collecting device can comprise or be, for example, a measuring tray as described in DE 10 2004 017 075 A1, other measuring trays, or another device suitable for holding, resting, or collecting fertilizer granules at or near the impact site. A combination of different collecting devices, for example, one or more measuring trays and one or more adhesive mats, can also be used.

[0010] The fertilizer granules are spread by a fertilizer spreader, e.g., a centrifugal fertilizer spreader or a pneumatic fertilizer spreader. This fertilizer spreader can be a centrifugal spreader. It can comprise at least one, typically two, centrifugal discs, wherein the centrifugal disc(s) can in particular be driven in rotation. Alternatively, the fertilizer spreader can also be a pneumatic spreader, which pneumatically applies fertilizer via at least one pneumatic conveying line with one or more associated impact plates. The fertilizer spreader can further comprise a storage container and a dosing unit. Typically, the fertilizer granules are introduced via the dosing unit to the application mechanism, e.g., onto a centrifugal disc or into the pneumatic system. For example, the fertilizer spreader can, for example,Each centrifugal disc or pneumatic conveying line must include a metering element, through which the fertilizer granules (the fertilizer) can be applied in adjustable quantities onto the centrifugal disc or into the pneumatic conveying line. In the case of a pneumatic fertilizer spreader, the metering element can, for example, be a section of a metering roller.

[0011] For example, the fertilizer spreader can comprise two centrifugal discs arranged next to one another transversely to the intended direction of travel, wherein each centrifugal disc can comprise one, two or more throwing vanes. For each centrifugal disc, an introduction system for the fertilizer granules can be arranged, which is designed to guide the fertilizer granules in an adjustable radial and / or concentric direction to a point on the centrifugal disc (feed point). The spreading pattern of the fertilizer spreader can be influenced by adjusting one or more fertilizer spreader parameters, in particular e.g. the vane position of the throwing vanes on the centrifugal discs, the disc diameter of the centrifugal discs, the effective length of the throwing vanes, the mounting height of the centrifugal spreader, the inclination of the centrifugal spreader and / or the inclination of the spreading mechanism and / or the centrifugal discs, the speed of the centrifugal discs and / or the feed point of the fertilizer granules on the centrifugal disc.

[0012] A camera with which an image of the at least one sprinkled collecting device can be captured can be, for example, a digital camera, in particular the camera of a mobile device, e.g., a cell phone or a tablet. When capturing an image, the collecting device can be captured in at least one camera image.

[0013] The fertilizer granules are located in the image. The fertilizer granules can be located on a mobile device, the on-board computer, or on a server. In particular, image analysis can be carried out, which includes a step that distinguishes fertilizer granules from the collection device, e.g. based on color differences, structure, deviation from a uniform pattern or a flat surface, and / or shadows. The actual distribution of the fertilizer granules on the collection device or along several collection devices can then be determined, for example by evaluating the area covered by fertilizer granules in comparison to the area visible from the base of the collection device and / or based on the known dimensions of the collection device.In particular, an actual distribution of the fertilizer granules on the collecting device can be calculated on one collecting device and / or along several collecting devices.

[0014] The enhancement step improves the localization of the fertilizer granules in the image. For example, such an enhancement step may include an improvement in image quality or image processing (compared to a method without such a step).

[0015] For example, the enhancement step may include detecting at least one circumstance parameter before capturing the image. A circumstance parameter may, in particular, describe one or more circumstances before and / or during capturing the image.

[0016] Such a circumstance parameter can, for example, include environmental information, information about the camera, information about the collecting device, information about the current image quality and / or further information. For example, the at least one circumstance parameter can include lighting conditions (in particular, e.g., brightness), a time of day, a shadow cast, in particular by a machine, e.g., the fertilizer spreader, the position of the sun, the current cloud cover, the direction, in particular with respect to the position of the camera, e.g.,in relation to a machine, in particular the fertilizer spreader, and / or the direction of the image recording, calculated (cast) shadows that fall on the collecting device or that are generated by the structure of the collecting device and the lighting conditions, a GPS position, the inclination of the ground and / or the collecting device, an attitude and / or orientation of the camera, a position (in particular distance to the target area) of the camera, an inclination of the camera, information about the shape of the collecting device, information about an identification or marking on the collecting device, information about the camera, e.g. model, manufacturer, active setting(s) of the camera, e.g. active filters, focal length, aperture, exposure time, sensitivity, resolution, information about a mobile device (which may, for example, comprise the camera) such as its operating system and operating system version, the recording software, the version of the recording software, and / or similar.Such a circumstance parameter can be taken into account for the setting when taking the image and / or during the evaluation.

[0017] Based on the at least one recorded circumstantial parameter, at least one setting parameter can be adjusted when capturing the image. For example, the exposure time and / or sensitivity can be adjusted to the lighting conditions, the desired recording direction to the cardinal direction, particularly with respect to the camera position, e.g., relative to a machine, e.g., the fertilizer spreader, and / or the direction of the image capture, and / or the position of the sun and / or the calculated shadow. Alternatively or additionally, it can be checked or configured that no filters and / or specific filters are applied.

[0018] For example, a predetermined resolution (optionally without a filter) can be used to capture an image. This can enable standardized evaluation and ensure compatibility when changing hardware. For example, the resolution can be FHD (Full High Definition - 1920x1080) and / or the optimal resolution of the camera can be used and downscaled to FHD (or another predetermined resolution). For example, a whole number of pixels can be combined, e.g., 2x2 or 3x3 pixels can be combined. This reduces the resolution to a fraction of the previous resolution, where the previous resolution is a multiple of the resulting resolution. This can be computationally advantageous and also increase accuracy, as the influence of pixel defects or similar can be reduced.If the previous resolution is reduced to a non-integer factor, pixels must be interpolated. This can, in particular, lead to blurring of sharp edges.

[0019] The at least one circumstance parameter can comprise an environmental parameter and / or a fertilizer spreader parameter. Example environmental parameters can include, for example, lighting conditions, position of the sun, cloud cover, direction, calculated shadows, e.g. calculated (cast) shadows falling on the collecting device, in particular due to the structure of the collecting device or a machine, e.g. the fertilizer spreader, and / or the like. Fertilizer spreader parameters can, for example, comprise one or more of the above-mentioned fertilizer spreader parameters, in particular the control setting(s), e.g. speed, throwing distance, working width, drop point, discharge angle, dosing rate, and / or property(ies) of the fertilizer, in particular, for example, grain size, average grain size, surface properties of the fertilizer grains, size distribution of the fertilizer grains, grain shape, true density, bulk density and / or grain strength.The fertilizer spreader parameters can be stored, for example (particularly in an app), e.g. in a mobile device, the on-board computer or on a server, which can optionally also be configured to carry out one or more further steps of the process.

[0020] For example, in the case of a fertilizer spreader, taking into account an environmental parameter as a setting parameter, the lighting of the fertilizer spreader, in particular a work light directed at the area or the spreading fan, in particular the collecting devices, can be switched on if this can contribute to an improvement in the image due to the lighting conditions.

[0021] The at least one circumstance parameter may in particular comprise a camera parameter, for example the camera setting, the type of filters activated in the camera, the white balance set on the camera, the exposure time, the focal length, the sensitivity, the aperture, the ISO setting, an image parameter (e.g. the sharpness and / or the resolution) and / or other (recording) settings.

[0022] For example, a camera setting parameter, e.g., an exposure time, the selected white balance, an ISO setting, focal length, aperture, a filter, a camera angle, tilt angle, and / or other camera settings, can be adjusted based on at least one circumstantial parameter, in particular a camera parameter, in particular set to a (new, calculated) target value. Such an adjustment of one (or more) setting parameters to a (new, calculated) target value can be carried out in particular based on at least one circumstantial parameter, e.g., a previously measured or known actual value.

[0023] For example, a white balance can be performed based on at least one circumstantial parameter, e.g., the measured value of a color chart or reference color in the image, which may be visible on the collection device, for example, and / or the color of the fertilizer. Thus, the camera's setting parameter that specifies the white balance can be adjusted appropriately for the image capture. Alternatively or additionally, lighting conditions and / or camera sensitivity can be used as circumstantial parameters to adjust one or more camera settings.

[0024] Optionally, for example, if the at least one circumstance parameter includes a camera parameter, instructions for improving the at least one camera parameter can be output. For example, the user can be instructed to change their position, e.g., if the user would otherwise create unfavorable shading and / or if the camera is held in an unfavorable manner (e.g., tilted).

[0025] If, for example, it is determined based on a camera parameter that the image is unsuitable (e.g. it is out of focus, distorted by excessive camera tilt, has strong shadows and / or is incorrectly exposed), this image can be rejected. The user can be instructed to improve at least one camera parameter, e.g. by changing the exposure time, changing the camera tilt and / or taking another image from a better position. Such instructions can be provided, for example, by guiding the user using optical and / or acoustic signals. For example, the user can be guided in the right direction by arrows on a mobile device, e.g. in the camera display of a mobile device, and / or instructed to improve the position by acoustic signals, e.g. a voice output or an alarm signal in the event of poor positioning.Alternatively or additionally, the user can be guided through augmented reality. For example, user guidance can be overlaid on a visible camera display image, e.g., to indicate a suitable inclination, position, orientation, and / or other camera parameters.

[0026] For example, instructions can be given to intentionally take a picture at an oblique angle to the collection device in order to avoid cast shadows. The camera parameter can in particular include (or be) a parameter that describes whether the edge of the collection device remains in the image. If it does not, the image cannot be cropped to the collection device, which means that information about the area of ​​the collection device may be lost. In such a case, the user can, for example, be instructed to change the image section. The guidance can, for example, be provided visually, by arrows or similar, by displaying images on the screen and / or by an audio guide. Optionally, the image can be taken automatically if the at least one camera parameter fulfills one or more specific conditions.For example, continuous or intermittent image recording (similar to a film recording or continuously recorded individual still images) can be performed, and an automatic triggering can occur when the catching device is correctly positioned in the image, for example, when it is sharp, complete, and optionally undistorted in the image (e.g., recognized as correctly positioned based on markings). In other embodiments, distortion can be accepted during automatic recording and later calculated out using the known geometry of the catching device.

[0027] Optionally or additionally, e.g. if a camera parameter detects that the image is out of focus, under certain other conditions or always, image stabilization can be performed, e.g. by an image stabilizer in hardware or software.

[0028] The enhancement step may include detecting a region of interest, for example, detecting when the collection device is completely within the image. Optionally, the image in which a region of interest has been detected may be cropped to that region. Such detection of a region of interest may, for example, occur before or after the image is captured. Thus, an image may be cropped to include a section of a collection device, so that it includes only the areas relevant for determining the actual distribution of the fertilizer granules in the resulting (cropped) image.

[0029] Furthermore, the enhancement step may additionally or alternatively comprise an image processing step that improves the differentiation of the fertilizer granules from the substrate. This enhancement step may incorporate all previously determined circumstantial parameters and other parameters, such as information about the fertilizer type and / or information about the collection device, e.g., the mat, or other information. For example, if the type, in particular the color of the fertilizer type, the color of the mat, and optionally other parameters, e.g., the white balance, are known, the fertilizer granules can be localized in an image based on these color distributions.

[0030] Optionally, the image processing step, which improves the discrimination of the fertilizer granules from the subsoil, may include a filtering step.

[0031] For example, the filtering step can comprise a convolution step. A convolution can, for example, comprise pixel-by-pixel iterating routines that process a partial environment of an image, in particular an environment of some pixels, e.g. all direct neighbors or a specific number of pixels around the pixel currently being processed, and can thus improve the differentiation of the fertilizer grains from the substrate, in particular can emphasize image characteristics relevant to the spreading pattern. A convolution step can, for example, be carried out using a convolution matrix (convolution kernel). With such a convolution step, for example, a sharpening (of some or all edges) of the image can be carried out and / or an accentuation of properties of the image and / or other properties of the image can be changed.

[0032] Alternatively or additionally, the filtering step may include a filtering step with a specialized filter designed and / or trained on existing images.

[0033] Additionally or alternatively, the filtering step may comprise a dilation step and / or an erosion step and / or a histogram adjustment (and / or a contrast spreading).

[0034] By means of a dilation step, structures in an image can be changed. In particular, existing structures can be highlighted, for example by means of a structured element. For example, fertilizer grains present in an image can be enlarged so that they can be better recognized during subsequent processing.

[0035] In the case of erosion, structures can be removed from an image, for example, using a mask. This can make the shape of existing structures, such as fertilizer grains, more easily recognizable, and potentially adjacent fertilizer grains can be more easily separated in the image.

[0036] A histogram can describe the brightness and / or contrast and / or color values ​​of an image. Histogram adjustment can adjust the brightness and / or contrast, e.g., to a reference histogram. For example, histogram adjustment can include evenly distributing the brightness across the image. This can lead to improved differentiation of fertilizer grains from the background, especially in unfavorable lighting conditions or shadows. Contrast spreading (distribution of the measured brightness values ​​across the entire possible range) can also lead to improved differentiation of fertilizer grains from the background.

[0037] The filtering step may include a fast Fourier transform and / or threshold filtering.

[0038] A fast Fourier transform, for example, can allow the image information to be viewed in the frequency domain. This can optionally be used, for example, with an additional filter, to filter out image information with a specific frequency, such as that generated by the structure of a collection device. In particular, a fast Fourier transform can be used to view only frequencies from the image that are lower than a regularly recurring structure of the collection device, so that only the (irregular) fertilizer distribution, but not a regular structure of the collection device, is detected.

[0039] Filtering with a threshold value can, for example, comprise filtering with a pixel's gray value, a specific pixel color value, or something similar. Such threshold filtering can be used, particularly if the color of the fertilizer and / or the color of the collection device is known, to distinguish whether a pixel is to be assigned to fertilizer or to the collection device.

[0040] The image processing step that improves the discrimination of the fertilizer grains from the substrate may in particular comprise one or more of the aforementioned steps, e.g. a convolution step with a fast Fourier transform and optionally further steps.

[0041] The enhancement step can be adapted taking into account the at least one circumstantial parameter. For example, one or more parameters included in the enhancement step can be reviewed based on the at least one circumstantial parameter and, if they do not correspond to the intended value, selected or redefined. Thus, the enhancement step can be adapted to one or more circumstances (during or) before the image is captured, which can, in particular, prepare for optimal image evaluation.

[0042] For example, an image processing step can be adapted taking into account the at least one circumstantial parameter, in particular, for example, the brightness, cloud cover, shadows, lighting conditions, fertilizer type, frequency and / or regularity of the structure of a collecting device, and / or a color of a collecting device. For example, the frequencies considered in a fast Fourier transform can be adapted taking into account the frequency of the structure of the collecting device, a filtering with a threshold value can be adapted to the brightness and / or lighting conditions, and / or a histogram adjustment can be adapted to the lighting conditions and / or shadows.

[0043] In sunlight and shadows, the enhancement step can include a histogram adjustment (especially locally) to create a uniform image, taking into account the exposure conditions, particularly those caused by sunlight and shadows. Alternatively or additionally, for small fertilizer grains, the enhancement step can include an adjustment of a Fourier filter to possibly still capture the small structures of the fertilizer grains. Alternatively or additionally, an erosion filter can be omitted from the enhancement step to avoid further reducing the size of the small fertilizer grains in the image.

[0044] At short focal lengths, the improvement step may include rectifying the image, as distortion is more pronounced at the edges.

[0045] For example, a class can be used to adapt the improvement step taking into account the at least one circumstance parameter. In particular, circumstance parameters taken into account in the improvement step and / or steps performed therein can be selected and / or combined based on the circumstances, which can be described in particular by the circumstance parameters.

[0046] The improvement step may include a training step for future steps of a method for determining an actual distribution of fertilizer grains. The training step may, in particular, serve to improve the method based on the experience gained. For example, the training step may involve training the step of capturing an image, the step of locating the fertilizer grains in the image, parts thereof, and / or intermediate steps. In particular, one or more image processing steps (for future image processing) may be trained in a training step.

[0047] In particular, a training step can be carried out, e.g. after completion of the calculation of an actual distribution of the fertilizer grains on the collecting device and / or along several collecting devices.

[0048] The training step can, for example, be performed by a class. The class can, for example, be the class of a mobile device, such as a smartphone or tablet.

[0049] In the training step, one or more intermediate results from one or more methods for determining the actual distribution and / or one or more previously calculated actual distributions of the fertilizer grains can be considered, e.g., as feedback. This training step can be performed, for example, after each calculation of an actual distribution of the fertilizer grains, after a certain number of calculations of an actual distribution of the fertilizer grains, after a certain period of time, at the user's request, and / or according to other criteria.

[0050] The training step can consider one or more user inputs. For example, the user can enter the amount of fertilizer applied and / or other information, which can be used to check the plausibility of the calculated actual distribution or similar. Alternatively or additionally, a user input can include a user evaluation, in particular as feedback. Such a user evaluation can be used to determine, in particular, whether the improvement step was successful, in particular, for example, whether further steps of an improvement step, e.g., one or more further training or compensation steps, should be performed.

[0051] The improvement step can comprise a compensation step for foreign bodies detected on the collecting device. For example, these can be identified in the image based on their structure, size and / or color. For example, structures on the collecting device that are considerably larger or considerably smaller than fertilizer granules can be identified as foreign bodies. Alternatively or additionally, structures on the collecting device that have a different color than the collecting device and / or the fertilizer granules can also be detected as foreign bodies. Alternatively or additionally, structures on the collecting device that have a different geometry than fertilizer granules can also be detected as foreign bodies. A different geometry can in particular comprise that the foreign bodies, e.g. plant residues or leaves, have a multiple of the average grain size of the set fertilizer in one dimension.Thus, for example, they may be ignored when locating the fertilizer granules in the image, for example by not taking them into account in the image processing after they have been identified as foreign bodies.

[0052] Typically, the analysis is pixel-by-pixel. In particular, the detected grains can be marked with a pixel-by-pixel mask. When foreign bodies are detected, all pixels associated with foreign bodies can then be removed from the pixel-by-pixel marking of fertilizer grains, leaving only the fertilizer grains.

[0053] Image processing can be parallelized. For example, during image processing, the image can be broken down into sub-images that are processed separately and optionally simultaneously, with the results then combined. For example, the image can be broken down into as many sub-images (or a multiple of this number) as the number of processor cores of the mobile device available for image processing, or a predetermined number of sub-images can be broken down into the number of images that can be processed in parallel on a server used for image processing.

[0054] An improvement step may further include guiding a user through the recording process. For example, a user may be guided visually and / or acoustically through the recording process. For example, a camera, such as a camera of a mobile device, may visually indicate (and / or acoustically output) the next steps the user should perform.

[0055] For example, the user can be guided to next take an image of at least one capture device. Based on environmental parameters, the user can be guided to take an image from a specific angle and / or with specific camera settings. The user can, for example, be instructed to change the recording position, for example if a previous image was blurred and / or distorted and / or based on the environmental parameters, in particular including the camera settings, a good image, in particular one suitable for evaluation, is not expected from the current position. This can be the case, for example, if the orientation of the camera towards the sun threatens overexposure or if the sun threatens to cast a shadow. Such guidance for the user can, for example, be provided by the display of arrows on a camera display or by acoustically outputting instructions or signal tones.A user can be alerted, for example visually or acoustically, if the section captured by the image is not suitable, for example only part of the catcher is shown, drop shadows are visible on the image and / or the image is distorted.

[0056] The user can also be guided to take a picture of a specific fall arrester next, particularly if more than one fall arrester is laid out in a field. After taking a picture of this fall arrester, the user can be guided to take a picture of the next capture device, whereby the order of the fall arresters to be captured can optionally be specified and optionally also changed. This can be particularly advantageous with complex deployment patterns of the capture devices. The user can, for example, be guided to capture several fall arresters (if present) in a specific row. This means that the order in which the fall arresters are to be captured can then be controlled and thus known.

[0057] During the subsequent evaluation, the position and / or identification of each collection device must be known so that the actual distribution of fertilizer grains can be calculated. When guiding the user, for example, the user interface of the camera and / or the mobile device can be used and instructions can be displayed on the user interface. The information can be output in the same setting as the camera user interface or in a different way, for example visually or acoustically as text. In particular, when guiding the user, instructions for improving a camera parameter as described above or another parameter, for example as described above for a camera parameter, can be displayed. The guidance can, for example, be provided visually, by arrows or similar, by displaying images on the screen and / or by an audio guide.

[0058] Optionally, the image can be captured automatically, e.g., if at least one camera or other parameter fulfills one or more specific conditions. For example, continuous image capture (similar to film recording or continuously recording individual still images) can be carried out and automatic triggering can occur when the capture device is correctly in the image, e.g., when it is sharp, complete, and optionally not distorted in the image (e.g., detected as correctly positioned based on markers). During continuous image capture, the user can be guided to change the position in a direction required for image capture, e.g., by moving or tilting the camera in a specific direction.

[0059] Likewise, based on an environmental parameter such as lighting conditions (e.g. brightness), the position of the sun or cloud cover, the recording mode can be adjusted, e.g. a flash can be switched on, external lighting can be switched on and / or a different evaluation routine (e.g. a different filter step, different white balance, different threshold values, additional steps such as contrast spreading or histogram adjustment and / or similar) can be used.

[0060] When calculating the actual distribution of the fertilizer granules, information about the collecting device can be taken into account. In particular, the geometry and / or the area and / or the color and / or a (regular) structure of the collecting device and / or image properties expected due to the (regular) structure of the collecting device, for example, shadows created by a structure of the collecting device, can be taken into account. In particular, one or more of these pieces of information about the collecting device can be taken into account during the image processing step, so that they are taken into account in the actual distribution of the fertilizer granules.

[0061] In particular, the known geometry of the catching device can be used to determine the camera tilt in the event of a distorted image. This camera tilt can be taken into account, especially during the image processing step.

[0062] For example, the camera tilt can be taken into account by determining the dimensions of the catching device based on the known geometry of the catching device and subtracting the camera tilt. Optionally, especially during the image processing step, only a region of interest within the catching surface can be considered, since the dimensions of the region under consideration can be derived from the rest of the catching device during image cropping. Alternatively or additionally, image areas with display defects, such as shadows, can be excluded from the evaluation.

[0063] The enhancement step may include pixel-by-pixel image segmentation using a neural network.

[0064] After capture, the image can be segmented pixel by pixel using a neural network, for example. For example, fertilizer grains can be highlighted using a neural network, making them easier to identify in further image analysis.

[0065] Additionally or alternatively, the improvement step can comprise classifying or marking detected objects using a neural network. For example, a neural network can be used to detect objects and to classify or mark them in the image. For example, the fertilizer grains and / or foreign bodies can be classified or marked, and further evaluation can be carried out on this basis. If, for example, the fertilizer grains have been classified or marked, counting the grains can be facilitated in the further process. Alternatively or additionally, an area (from the classification or the pixel-wise segmentation) can be used for evaluation. The improvement step can further comprise filtering using a specially trained neural network. Using such a trained neural network, for example, with the help of one or more circumstance parameters, such asBased on lighting or white balance information, an optimized filter can be determined and applied to the images. This can lead to simpler and better image evaluation.

[0066] The invention further comprises a system for capturing an image in a method for determining an actual distribution of fertilizer granules as described above. The system comprises, in particular, a screen (display), a processor, and a memory. The memory contains instructions that, when executed by a processor, visually and / or acoustically guide a user to capture an image of the at least one sprinkled collecting device. The guidance can be provided, for example, as described above.

[0067] The system's memory may further optionally include instructions that, when executed by a processor, perform the steps of locating the fertilizer granules in the image and calculating an actual distribution of the fertilizer granules on the collection device and / or along a plurality of collection devices, wherein at least one improvement step improves the localization of the fertilizer granules in the image. Such an improvement step of locating the fertilizer granules in the image may, in particular, comprise one or more of the improvement steps and measures described above.

[0068] The system may further comprise a communication means of the imaging device that may allow communication with an on-board computer, for example, a wireless communication module or a connection for a cable. Accordingly, the on-board computer may also comprise a communication means suitable for connection to the imaging device, for example, a wireless communication module (matching the communication module of the imaging device) or a connection for a cable (matching the connection of the imaging device). The system may optionally comprise one or more collecting devices.

[0069] If the actual distribution of fertilizer granules on more than one collection device is to be determined, the system can recommend or adjust an order in which the images of the collection devices should be taken.

[0070] Particularly with visual user guidance, the screen display can be adjusted to one or more environmental parameters. For example, based on an environmental parameter such as brightness, the position of the sun, or cloud cover, the screen display can be adjusted, e.g., by increasing the brightness or activating a night display mode.

[0071] Aspects of the above invention emerge from the attached figures, which are not to scale.

[0072] Figure 1 shows exemplary steps of a method in which a circumstance parameter is recorded,

[0073] Figure 2 shows exemplary steps of a process in which fertilizer granules are located,

[0074] Figure 3 exemplary steps of a process with an improvement step,

[0075] Figure 4 exemplary steps of a method with a training step,

[0076] Figure 5 shows an example user interface.

[0077] Figure 1 shows exemplary steps of a method. In particular, Figure 1 shows that at least one circumstance parameter can be recorded before or during the capture of the image (step 101). For example, the brightness can be recorded before the image is captured. Based on the at least one circumstance parameter, one or more setting parameters can be adjusted when the image is captured (step 102). For example, the exposure time can be adjusted to the brightness. The image can then be captured with the adjusted setting parameter. This can improve the localization of the fertilizer grains in the image (step 103) and a subsequent calculation of an actual distribution of the fertilizer grains on the collecting device and / or along several collecting devices (step 104).

[0078] Figure 2 shows exemplary steps of a method. In the example shown, an image is captured (step 201). Subsequently, a region of interest is identified in the image (step 202). The image can then be cropped to the region of interest (step 203). The fertilizer granules can then be located in the (cropped) image (step 204).

[0079] In other examples (not shown), the detection of a region of interest may already be performed before the image is captured. For example, a region of interest may be detected during continuous image capture (similar to film recording or continuously capturing individual still images). An automatic trigger (or a signal to the user) may then occur, for example, when the region of interest has been detected (e.g., the entire collection device) and is correctly in the image, for example, when it is in focus and completely in the image (e.g., detected as correctly positioned based on markers or an expected shape of the mat in the image).

[0080] Figure 3 shows exemplary steps of a method beginning with the acquisition of an image (step 301). After the acquisition of an image, the enhancement step is adapted taking into account at least one circumstantial parameter (step 302). In other examples (not shown here), the enhancement step can be adapted taking into account at least one circumstantial parameter before or during the acquisition of the image.

[0081] For example, the image processing step, e.g., filter selection, can be performed taking into account at least one circumstantial parameter. For example, a filter used, e.g., a threshold value, can be adjusted based on brightness. Alternatively or additionally, a filter, e.g., a threshold value, can be adjusted based on one or more circumstantial parameters related to the fertilizer and / or the collection device, e.g., the color of the fertilizer and / or the collection device.

[0082] After adapting the enhancement step taking into account at least one circumstance parameter, the enhancement step (step 303) can be performed, and an actual distribution of the fertilizer granules can be calculated (step 304). The enhancement step, which improves the localization of the fertilizer granules in the image, can optionally be included as part of the acquisition of an image and / or as part of the localization of the fertilizer granules in the image, e.g., during image processing. For example, an enhancement step, which includes a filtering step, can be included as part of the image processing, in particular, be part of the step of localizing the fertilizer granules in the image.

[0083] Figure 4 shows further exemplary method steps. In particular, Figure 4 shows a training step (step 402), particularly for future steps of a method for determining an actual distribution of fertilizer grains. Such a training step (step 402) can be performed, for example, after calculating an actual distribution of the fertilizer grains on the collecting device and / or along multiple collecting devices (401).

[0084] Such a training step can, for example, consider the actual distribution of the fertilizer grains, user inputs, and / or one or more intermediate results from a method for determining the actual distribution. It can, for example, be performed using a classifier. The training step for future image processing steps can thus, in particular, improve a method for determining the actual distribution, in particular the localization of the fertilizer grains in the image, and make the resulting results more accurate.

[0085] Figure 5 shows an exemplary user interface as it can be used in a method for determining an actual distribution of fertilizer grains or a system for capturing an image in a method for determining an actual distribution of fertilizer grains.

[0086] The user interface can be displayed, in particular, on a display 1, e.g., the display of a mobile device. In the example shown, a collecting device 2 and fertilizer granules 4 are visible on and next to the collecting device 2. The collecting device is not centered in the image of the display 1, and the user is visually instructed on the user interface, here, for example, by means of arrow 3a and an instruction 3b, to move the camera so that the collecting device is centered.

[0087] An image can then be captured. In other embodiments, the user guidance may additionally or alternatively include augmented reality and / or non-visual components, e.g., acoustic components such as a warning signal or other components, e.g., a vibration alarm.

Claims

Claims Method for determining an actual distribution of fertilizer grains comprising the steps of: a) laying out at least one collecting device for fertilizer grains; b) spreading the fertilizer grains over the at least one collecting device using a centrifugal fertilizer spreader; c) taking an image of the at least one sprinkled collecting device with a camera, d) locating the fertilizer grains in the image, and e) calculating an actual distribution of the fertilizer grains on the collecting device and / or along a plurality of collecting devices, wherein the method comprises at least one improvement step that improves the localization of the fertilizer grains in the image. Method according to claim 1, wherein the improvement step comprises detecting at least one circumstance parameter before capturing the image. Method according to claim 2, wherein at least one setting parameter is adjusted during the capturing of the image based on the at least one detected circumstance parameter.The method according to claim 2 or 3, wherein the at least one circumstance parameter comprises an environmental parameter and / or a fertilizer spreader parameter. The method according to claim 2 or 3, wherein the at least one circumstance parameter comprises a camera parameter, optionally outputting guidance for improving the at least one camera parameter and / or. wherein the image is automatically captured when the at least one camera parameter meets certain conditions. Method according to claims 1 to 5, wherein the enhancing step comprises detecting a region of interest, wherein the image is optionally cropped to the region of interest. Method according to claims 1 to 6, wherein the enhancing step comprises an image processing step that improves the discrimination of the fertilizer grains from the subsoil. Method according to claim 7, wherein the image processing step that improves the discrimination of the fertilizer grains from the subsoil comprises a filtering step, wherein optionally the filtering step comprises a convolution step and / or wherein optionally the filtering step comprises a dilation step and / or an erosion step and / or a histogram fitting and / or wherein the filtering step comprises a fast Fourier transform and / or filtering with a threshold value.The method according to claims 2 to 8, wherein the improvement step is adapted taking into account the at least one circumstance parameter and / or wherein the improvement step comprises a training step for future steps of a method for determining an actual distribution of fertilizer granules. The method according to claims 1 to 9, wherein the improvement step comprises a compensation step for foreign bodies detected on the collecting device. The method according to claims 1 to 10, wherein the image processing is parallelized. Method according to claims 1 to 11, wherein the improvement step comprises guiding a user through the recording process, in particular being guided visually and / or acoustically through the recording process. Method according to claims 1 to 12, wherein information about the collecting device is taken into account when calculating the actual distribution of the fertilizer grains. Method according to claims 1 to 13, wherein the improvement step comprises pixel-by-pixel image segmentation using a neural network and / or wherein the improvement step comprises classifying or marking recognized objects using a neural network. Method according to claims 1 to 14, wherein the improvement step comprises filtering using a specially trained neural network.System for capturing an image in a method for determining an actual distribution of fertilizer grains, characterized in that the system comprises a screen, a processor and a memory which comprises instructions which, when executed by a processor, optically and / or acoustically guide a user to capturing an image of the at least one sprinkled collecting device, wherein the memory optionally comprises instructions which, when executed by a processor, after image capture, carry out the steps. Locating the fertilizer granules in the image, and Calculate an actual distribution of the fertilizer granules on the collecting device and / or along a plurality of collecting devices, wherein at least one improvement step improves the localization of the fertilizer granules in the image.

Images