阅读说明：本技术 用于杂草控制的装置 (Device for weed control ) 是由 H.巴斯费尔德 T.阿里安斯 P.戴 V.吉鲁 J.哈德洛 于 2018-07-02 设计创作，主要内容包括：本发明涉及用于杂草控制的装置。描述了给处理单元提供(210)环境的至少一张图像。所述处理单元分析(220)所述至少一张图像以从多个植被控制技术中确定要用于所述环境的至少第一部分的杂草控制的至少一个植被控制技术。输出单元输出(230)可用于激活所述至少一个植被控制技术的信息。(The present invention relates to a device for weed control. Providing (210) at least one image of an environment to a processing unit is described. The processing unit analyzes (220) the at least one image to determine at least one vegetation control technique from a plurality of vegetation control techniques to be used for weed control of at least a first portion of the environment. An output unit outputs (230) information usable to activate the at least one vegetation control technique.)

1. A device (10) for weed control comprising:

-an input unit (20);

-a processing unit (30); and

-an output unit (40);

wherein the input unit is configured to provide the processing unit with at least one image of an environment;

wherein the processing unit is configured to analyze the at least one image to determine at least one vegetation control technique from a plurality of vegetation control techniques to be used for weed control of at least a first portion of the environment; and is

Wherein the output unit is configured to output information usable to activate the at least one vegetation control technique.

2. The apparatus of claim 1, wherein analyzing the at least one image to determine at least one vegetation control technique comprises determining at least one location of vegetation in the at least first portion of the environment, and wherein the processing unit is configured to determine the at least one vegetation control technique to be used at the at least one location.

3. The apparatus according to any of claims 1-2, wherein the at least one image is obtained by at least one camera, and wherein the input unit is configured to provide the processing unit with at least one location associated with the at least one camera at the time the at least one image was obtained.

4. The apparatus of any of claims 1-3, wherein analyzing the at least one image to determine the at least one vegetation control technique comprises determining at least one type of weed.

5. The apparatus of claim 4, wherein the processing unit is configured to determine at least one location of the at least one type of weed.

6. The apparatus of any of claims 1-5, wherein analyzing the at least one image to determine the at least one vegetation control technique comprises determining a first type of weed in the at least first portion of the environment and determining a second type of weed in at least a second portion of the environment.

7. The apparatus of claim 6, wherein the processing unit is configured to analyze the at least one image to determine a first vegetation control technique to be used for weed control of the first type of weed in the at least first portion of the environment; and wherein the processing unit is configured to analyze the at least one image to determine a second vegetation control technique to be used for weed control of the second type of weed in at least a second portion of the environment.

8. The apparatus of any of claims 1-7, wherein the processing unit is configured to analyze the at least one image to determine a first vegetation control technique from the plurality of vegetation control techniques to be used for weed control of at least the first portion of the environment; and wherein the processing unit is configured to analyze the at least one image to determine a second vegetation control technique from the plurality of vegetation control techniques to be used for weed control of at least a second portion of the environment.

9. The apparatus of any of claims 1-8, wherein analyzing the at least one image comprises using a machine learning algorithm.

10. A system (100) for weed control, comprising:

-at least one camera (110);

-a device (10) for weed control according to any of claims 1-9; and

-at least one vegetation control technique (120);

wherein the at least one camera is configured to obtain the at least one image of the environment;

wherein the at least one vegetation control technology is mounted on a vehicle (130); and

wherein the device is configured to activate the at least one vegetation control technique for the at least a first portion of the environment.

11. The system of claim 10, wherein the device is mounted on the vehicle; and wherein the at least one camera is mounted on the vehicle.

12. A method (200) for weed control, comprising:

(a) providing (210) at least one image of the environment to the processing unit;

(c) analyzing (220), by the processing unit, the at least one image to determine at least one vegetation control technique from a plurality of vegetation control techniques to be used for weed control of at least a first portion of the environment; and

(e) information usable to activate the at least one vegetation control technique is output (230) by an output unit.

13. The method of claim 12, wherein step (c) includes the step of determining (240) at least one location of vegetation in the at least first portion of the environment; and wherein the method comprises the step (d) of determining (250), by the processing unit, the at least one vegetation control technique to be used at the at least one site.

14. The method according to any one of claims 12-13, wherein in step (a) the at least one image is obtained by at least one camera; and wherein the method comprises the step (b) of providing (260) the processing unit with at least one location associated with the at least one camera when obtaining the at least one image.

15. A computer program element for controlling an apparatus according to any one of claims 1 to 9 and/or a system according to any one of claims 10 to 11, which, when being executed by a processor, is configured to carry out the method according to any one of claims 12 to 14.

Technical Field

The present invention relates to an arrangement for weed control, a system for weed control, a method for weed control, as well as a computer program element and a computer readable medium.

Background

The general background of the invention is weed control. Certain industrial areas and areas around railroad tracks require control of vegetation. For railways, this control improves visibility from the point of view of the people on the train (e.g. the driver), and also from the point of view of the people working on the track. Such control can lead to improved safety. Moreover, vegetation can interfere with or damage the track and associated signal and communication lines. Control of vegetation is required to mitigate this phenomenon. Vegetation control, also known as weed control, can be very good, time consuming and resource consuming, especially if performed manually. The weed spray train carries a herbicide contained in a chemical tank on the vehicle that can be sprayed onto the track and surrounding area to control vegetation. However, such weed control is expensive and there is an increasing desire among the general public to see a reduction in environmental impact.

Disclosure of Invention

It would be advantageous to have an improved means for weed control.

The object of the invention is achieved by the subject matter of the independent claims, wherein further embodiments are comprised in the dependent claims. It should be noted that the aspects and examples of the invention described below also apply to the apparatus for weed control, to the system for weed control, to the method for weed control, and to the computer program element and the computer readable medium.

According to a first aspect, there is provided an apparatus for weed control comprising:

-an input unit;

-a processing unit; and

-an output unit.

The input unit is configured to provide at least one image of the environment to the processing unit. The processing unit is configured to analyze the at least one image to determine at least one vegetation control technique from a plurality of vegetation control techniques for weed control of at least a first portion of the environment. An output unit is configured to output information usable to activate the at least one vegetation control technique. In other words, one or more images of the environment have been obtained. There are several possible vegetation control techniques available for weed control. The facility then analyzes the one or more images to determine which one or more of the available vegetation control techniques should be used to control weeds at one or more specific locations of the environment.

In this way, the most suitable vegetation control techniques can be used in different areas of the environment. Also, in different areas of the environment, different techniques may be used, each of which is most appropriate for each different area.

In this way, herbicide-based technologies, such as those that can be applied in the form of a spray, can be used only if they are the most suitable technologies for one or more specific areas of the environment. This also means that non-herbicidal technology is used at other areas of the environment. Thus, not only is overall control of the weeds improved, but the use of herbicides, and especially the most aggressive ones, is reduced because the most suitable technique is used for each area.

In an example, analyzing the at least one image to determine the at least one vegetation control technique includes determining at least one location of vegetation in at least a first portion of the environment. The processing unit is configured to then determine at least one of the vegetation control techniques to be used at that at least one location.

In other words, image processing may be used to determine the vegetation area in the obtained image from which the most appropriate technique to be used for weed control for that vegetation area may be selected. Also, the vegetation control technique may be applied only at the location of the vegetation, where the most suitable vegetation control technique may be used for each vegetation location.

In this way, the most suitable vegetation control technique can be selected for different vegetation areas, where small vegetation areas can be controlled by different means than large vegetation areas, for example.

In an example, the at least one image is obtained by at least one camera. The input unit is configured to provide the processing unit with at least one location associated with the at least one camera at a time when the at least one image was obtained.

The location may be a geographical location, an exact location relative to the ground, or a location on the ground that references a location of at least one vegetation control technique. In other words, an absolute geographical location may be utilized or an above-ground location may not necessarily be known in an absolute sense, but rather a location referenced to the location of the weed control technique.

Thus, by associating an image with the location at which the image was obtained, which may be an absolute geographical location or a location on the ground whose position relative to the vegetation control technique is known, the vegetation control technique can be accurately applied to the location.

In an example, analyzing the at least one image to determine the at least one vegetation control technique includes determining at least one type of weed. In other words, one or more types of weeds to be controlled may be considered in selecting an appropriate vegetation control technique. In an example, the processing unit is configured to determine at least one location of at least one type of weed. In other words, image processing can be used to determine the weed type and its location. The location may be a location in the image. The location may be a real geographic location. The location may be within the image and can be referenced to the location of one or more vegetation control techniques. In this way, by determining the location of a particular type of weed, an optimal vegetation control technique can be applied to that particular location, which also applies to different weeds requiring application of different vegetation control techniques at different locations.

In an example, analyzing the at least one image to determine the at least one vegetation control technique includes determining a first type of weed in at least a first portion of the environment and determining a second type of weed in at least a second portion of the environment.

The most suitable mode of operation for vegetation control technology can therefore be determined based on the different weed types in the environment. In an example, the processing unit is configured to analyze the at least one image to determine a first vegetation control technique to be used for weed control of a first type of weed in at least a first portion of the environment. The processing unit is configured to also analyze the at least one image to determine a second vegetation control technique to be used for weed control of a second type of weed in at least a second portion of the environment.

In other words, the most suitable vegetation control technique may be selected according to the particular type of weeds that will be found in the portion of the environment, thereby enabling the particular vegetation control technique to be applied only at the locations where those particular weeds will be found. In an example, the processing unit is configured to analyze the at least one image to determine a first vegetation control technique to be used for weed control of at least a first portion of the environment from a plurality of vegetation control techniques. The processing unit is configured to also analyze the at least one image to determine a second vegetation control technique from the plurality of vegetation control techniques to be used for weed control of at least a second portion of the environment.

In other words, a first technique may be selected for weed control at a first location of an environment based on image analysis, and a different vegetation control technique may be selected for weed control at a different location based on image analysis.

In this way, the most suitable vegetation control technique can be selected for a particular part of the environment, e.g., one weed control technique is used for some weeds while a different vegetation control technique is used for different weeds, and/or one planting control technique can be used for certain types of weeds in a first part of the environment while a different vegetation control technique is used for the same weeds in a different part of the environment. For example, the selected vegetation control technique may take into account the terrain on the ground, for example whether the terrain is dry, sandy, swamp, wet, or an area of particular environmental importance (a protected area) and these types of terrain are taken into account when selecting the most appropriate vegetation control technique for the same type (or different types) of weeds.

In another example, the most suitable vegetation control technique can be determined based on the stage of growth or stage of development of the weed species. According to a licensed embodiment, the development stage may be defined by BBCH (an internationally accepted code from biologische bundesanstalt, bundesportenamt und Chemische Industrie, germany).

In another example, the most suitable vegetation control technique can be determined based on the amount of weeds in the environment.

In addition, this means that chemically aggressive weed control means can be kept to a minimum.

In an example, analyzing the at least one image includes using a machine learning algorithm.

In the discussion above, and in the discussion below, vegetation control techniques may be referred to as weed control techniques, and vice versa.

According to a second aspect, there is provided a system for weed control comprising:

-at least one camera;

-a device for weed control according to the first aspect; and

-at least one vegetation control technique.

The at least one camera is configured to obtain at least one image of the environment. The at least one vegetation control technology is mounted on the vehicle. The apparatus is configured to activate the at least one vegetation control technique for at least a first portion of the environment.

In this manner, the vehicle may be moved around an environment and control weeds within the environment using different vegetation control techniques, where a particular vegetation control technique is determined based on an image of that environment. In this way, images are obtained by a platform, such as one or more drones flying over the environment. This information is sent to a device, which may be located in an office. The apparatus determines where in the environment what vegetation control technique should be used. This information can be provided within a weed control map, which is provided to a vehicle that moves around the environment and activates the required vegetation control techniques at specific parts of the environment.

In an example, the device is mounted on a vehicle. In an example, at least one camera is mounted on the vehicle.

According to a third aspect, there is provided a method for weed control comprising:

(a) providing at least one image of the environment to a processing unit;

(c) analyzing, by the processing unit, the at least one image to determine at least one vegetation control technique from a plurality of vegetation control techniques to be used for weed control of at least a first portion of the environment; and

(e) information is output by the output unit, the information being usable to activate the at least one vegetation control technique.

In an example, step (c) includes the step of determining (240) at least one vegetation location in at least a first portion of the environment; and wherein the method comprises the step (d) of determining (250), by the processing unit, at least one vegetation control technique to be used at that at least one location.

In an example, the at least one image in step (a) is obtained by at least one camera; and wherein the method comprises providing (260) the processing unit with at least one location related to the at least one camera at the time the at least one image was obtained.

According to a further aspect, a computer program element for controlling an apparatus according to the first aspect and/or a system according to the second aspect is provided, which program element, when being executed by a processor, is configured to carry out the method of the third aspect. Advantageously, the benefits provided by any of the above aspects apply equally to all other aspects, and vice versa.

The above aspects and examples can be understood and appreciated with reference to the embodiments described below.

Drawings

Exemplary embodiments will now be described with reference to the following drawings:

figure 1 shows a schematic arrangement of an example of an arrangement for weed control;

FIG. 2 shows a schematic arrangement of an example of a system for weed control;

FIG. 3 illustrates a method for weed control;

FIG. 4 shows a schematic arrangement of an example of a system for weed control;

FIG. 5 shows a schematic arrangement of an example of a system for weed control;

FIG. 6 shows a schematic arrangement of an example of a part of a system for weed control;

FIG. 7 shows a schematic arrangement of an example of a part of a system for weed control;

FIG. 8 shows a schematic arrangement of an example of a part of a system for weed control;

FIG. 9 shows a schematic arrangement of a portion of the system for weed control shown in FIG. 7 in more detail; and

fig. 10 shows a schematic depiction of a railway track and surrounding area.

Detailed Description

Fig. 1 shows an example of an apparatus 10 for weed control. The apparatus 10 comprises an input unit 20, a processing unit 30 and an output unit 40. The input unit 20 is configured to provide the processing unit 30 with at least one image of the environment. The processing unit 30 is configured to analyze the at least one image to determine at least one vegetation control technique to be used for weed control of at least a first portion of the environment from a plurality of vegetation control techniques. The output unit 40 is configured to output information that can be used to activate the at least one vegetation control technique. The at least one vegetation control technique is activated at least a first portion of the environment.

In an example, the plant is operated in real time, where images are obtained and immediately processed, and the determined vegetation control technique is immediately used to control weeds. Thus, for example, a vehicle may obtain an image of its environment and process the image to determine which vegetation control techniques carried by the vehicle should be used for a particular portion of its environment.

In an example, the plant is operated in near real-time, where an image of the environment is obtained and the image is immediately processed to determine which vegetation control technique should be used to control weeds at a particular area of that environment. This information can then be used by one or more suitable systems traveling in the environment and apply suitable vegetation control techniques to particular portions of the environment. Thus, for example, a first vehicle equipped with one or more cameras, such as a car, train, truck, or Unmanned Aerial Vehicle (UAV) or drone, may travel within the environment and obtain images. This image can be immediately processed to determine a "weed map" detailing where in the environment a particular vegetation control technique should be used. Thereafter, a transport vehicle equipped with several different vegetation control technologies can travel through the environment and apply specific identified weed control technologies to different specific areas of the environment. In another example, several different vehicles are each equipped with a single vegetation control technology, travel in the environment and use their specific vegetation control technology only for those specific areas of the environment where it has been determined that vegetation control technology should be used.

In an example, the apparatus operates in an offline mode. Thus, the image that has been obtained before is supplied to the apparatus later. The setup then determines where in an area a particular vegetation control technique should be used and actually generates a weed map. The weed map is then later used by one or more vehicles that then travel in the area and apply specific vegetation control techniques to specific parts of the environment.

In an example, the output unit outputs a signal that can be directly used to activate vegetation control techniques.

According to an example, analyzing the at least one image to determine the at least one vegetation control technique includes determining at least one location of vegetation in at least a first portion of the environment. The processing unit is configured to then determine at least one vegetation control technique to be used at that at least one site.

According to an example, the at least one image is obtained by at least one camera. The input unit is configured to then provide the processing unit with at least one location associated with the at least one camera at the time the at least one image was obtained.

In an example, the place is an absolute geographic place.

In an example, the location is a location determined with reference to a location of at least one vegetation control technique. In other words, it may be determined that the image relates to a particular location on the ground without knowing its exact geographical location, but by knowing the position of the at least one vegetation control technique relative to that location at the time the image was obtained, the vegetation control technique may then be applied at that location at a later time by moving the vegetation control technique to that location.

In an example, a GPS unit is used to determine the location of at least one camera at the time a particular image is obtained, and/or is used in determining the location.

In an example, an inertial navigation unit is used alone, or in combination with a GPS unit, to determine the location of at least one camera at the time a particular image is obtained. Thus, for example, an inertial navigation unit comprising, for example, one or more laser gyroscopes, is calibrated or nulled at a known location, and as it moves with at least one camera, movement in x, y, z coordinates away from that known location can be determined, from which movement the location of the at least one camera at the time the image was obtained can be determined.

In an example, image processing of the obtained images is used alone, or in combination with a GPS unit and an inertial navigation unit, to determine the location of at least one camera at the time a particular image was obtained. Thus, the visual markers may be used individually or in combination to determine the location of the camera. According to an example, analyzing the at least one image to determine the at least one vegetation control technique includes determining at least one type of weed.

According to an example, the processing unit is configured to determine at least one location of the at least one type of weed.

According to an example, analyzing the at least one image to determine the at least one vegetation control technique includes determining a first type of weed in at least a first portion of the environment and determining a second type of weed in at least a second portion of the environment.

According to an example, the processing unit is configured to analyze the at least one image to determine a first vegetation control technique to be used for weed control of a first type of weed in at least a first portion of the environment. The processing unit is configured to further analyze the at least one image to determine a second vegetation control technique to be used for weed control of a second type of weed within at least a second portion of the environment.

According to an example, the processing unit is configured to analyze the at least one image to determine a first vegetation control technique to be used for weed control of at least a first portion of the environment from a plurality of vegetation control techniques. The processing unit is configured to also analyze the at least one image to determine a second vegetation control technique to be used for weed control of at least a second portion of the environment from the plurality of vegetation control techniques.

In an example, the at least a second portion of the environment is different from the at least a first portion of the environment.

Thus, it can be determined that different weeds are located in different parts of the environment to enable the determination of the most suitable vegetation control technique for those areas. In an example, the at least a second portion of the environment is at least partially bounded by the at least a first portion of the environment.

In other words, one area of the environment is found to be located within another area of the environment. One vegetation control technique may be used for a large area at this time, while another vegetation control technique may be used for a smaller area that would be found in that area.

In an example, the at least a second portion of the environment is at least a subset of the at least a first portion of the environment.

Thus, for example, a smaller area of a particular type of weed may be found within a larger area of the weed. For example, one or more dandelion may be located within a grass region. At this point, a first vegetation control technique may be used throughout the grass field, including where the dandelion are located. This vegetation control technique may be selected as the technique suitable for controlling grass, and is not necessarily the most aggressive vegetation control technique available. However, where a more difficult to kill weed such as dandelion is found for that subset of the grass field, a more aggressive vegetation control technique, such as chemical spraying at that particular site, may be used. In this way, the amount of chemical spray can be minimized.

According to an example, analyzing the at least one image includes utilizing a machine learning algorithm.

In an example, the machine learning algorithm comprises a decision tree algorithm.

In an example, the machine learning algorithm includes an artificial neural network.

In an example, a mechanical learning algorithm is taught on the basis of a plurality of images. In an example, a machine learning algorithm is taught based on a plurality of images including images of at least one type of weed. In an example, a machine learning algorithm is taught based on a plurality of images including images of a plurality of weeds.

In an example, the at least one vegetation control technique includes one or more of: one or more chemicals; chemical spraying; a chemical liquid; a chemical solid; high-pressure water; high-temperature water; high-temperature high-pressure water; steam; electrical energy; electric induction; current flow; high voltage energy; electromagnetic radiation; x-ray radiation; ultraviolet radiation; visible radiation; microwave radiation; pulsed laser radiation; a flame system.

Fig. 2 shows an example of a system 100 for weed control. The system 100 includes at least one camera 110, and a device 10 for weed control as described in any of the examples above in connection with fig. 1. The system 100 also includes at least one vegetation control technique 120. The at least one camera 110 is configured to obtain at least one image of the environment. The at least one vegetation control technique 120 is mounted on a vehicle 130. The apparatus 10 is configured to activate the at least one vegetation control technique 120 for at least a first portion of the environment.

In the example, the apparatus 10 is mounted on a vehicle 130. In an example, at least one camera 110 is mounted on the vehicle 130.

In an example, the vehicle is a train, or a railway wagon, wagon or truck, or a uni mug.

In an example, the input unit is configured to provide the processing unit with at least one geographical location associated with the at least one camera at the time the at least one image was obtained.

In an example, the apparatus is configured to activate the at least one vegetation control technique based on at least one geographic location associated with the at least one camera and a spatial relationship between the at least one camera and the at least one vegetation control technique at the time the at least one image was obtained. In this way, by knowing where the image was obtained by the camera mounted on the vehicle and also knowing where the vegetation control technique is mounted on the vehicle relative to the camera, forward movement of the vehicle can be taken into account to activate the vegetation control technique at the same place where the image was obtained, and indeed within the imaging area.

In an example, the apparatus is configured to activate the first vegetation control technique before activating the second vegetation control technique or to activate the first vegetation control technique after activating the second vegetation control technique.

In an example, the first vegetation control technology is mounted ahead of the second vegetation control technology relative to the direction of travel of the vehicle or the first vegetation control technology is mounted behind the second vegetation control technology relative to the direction of travel of the vehicle.

Fig. 3 shows the basic steps of a method 200 for weed control. The method 200 comprises the following steps:

in a providing step 210, also referred to as step (a), at least one image of the environment is provided 210 to the processing unit 30;

in an analyzing step 220, also referred to as step (c), the at least one image is analyzed by the processing unit to determine at least one vegetation control technique from a plurality of vegetation control techniques to be used for weed control of at least a first portion of the environment; and is

In an output step 230, also referred to as step (e), information is output by the output unit 40 that can be used to activate the at least one vegetation control technique.

In an example, at least one image of the environment is provided from the input unit 20 to the processing unit.

According to an example, step (c) includes the step of determining 240 at least one location of vegetation in at least a first portion of the environment. The method then includes the step (d) of determining 250, by the processing unit, the at least one vegetation control technique to be used at the at least one site.

According to an example, in step (a) at least one image is obtained by at least one camera, and the method comprises a step (b) of providing 260 to the processing unit at least one location associated with the at least one camera when the at least one image was obtained.

In an example, step (c) includes determining 270 at least one type of weed.

In an example, step (c) includes determining 280 at least one location of the at least one type of weed.

In an example, step (c) includes determining 290 a first type of weed in at least a first portion of the environment and determining 300 a second type of weed in at least a second portion of the environment.

In an example, step (c) includes determining 310 a first vegetation control technique to be used for weed control of a first type of weed in at least a first portion of the environment, and determining 320 a second vegetation control technique to be used for weed control of a second type of weed in at least a second portion of the environment.

In an example, step (c) includes determining 330 a first vegetation control technique to be used for weed control of at least a first portion of the environment; and determining 340 a second vegetation control technique to be used for weed control of at least a second portion of the environment.

In an example, the at least a second portion of the environment is different from the at least a first portion of the environment.

In an example, the at least a second portion of the environment is at least partially bounded by the at least a first portion of the environment.

In an example, the at least a second portion of the environment is at least a subset of the at least a first portion of the environment.

In an example, step (c) includes utilizing 350 a machine learning algorithm.

In an example, a method includes using a vehicle, and wherein the method includes obtaining at least one image of an environment by at least one camera; and activating the at least one vegetation control technology installed on the vehicle for at least a first portion of the environment.

In an example, a method includes mounting a processing unit, an output unit, and at least one camera on a vehicle.

In an example, the method includes activating the first vegetation control technique before activating the second vegetation control technique or activating the first vegetation control technique after activating the second vegetation control technique.

The apparatus, system and method for weed control, which relate to weed control in a railway track environment, is now described in more detail in connection with fig. 4-10, with the weed control technology being installed on various parts of the train.

Fig. 4 shows an example of a system 100 for weed control. Several drones have cameras 110. These drones fly along the railway track. The cameras obtain images of the environment of the railroad track, which are the ground between the tracks and the ground on both sides of the tracks. The environment being imaged is the environment required for weed control. There need not be several drones, but one drone with one camera 110 can obtain the necessary images. In practice, the images may be obtained by one or more cameras 110 held by a person accessing the railway track environment, by an airplane, or by one or more cameras 110 held by, for example, a train traveling along the railway track. The resolution of the image obtained by the camera 110 enables the vegetation to be identified as vegetation and indeed may be at a resolution that can distinguish one type of weed from another. The obtained image may be a color image but need not be. The images obtained by the drone are transmitted to the device 10. The image may be transmitted to the device 10 as it is acquired by the camera 110, or may be transmitted after it is acquired, for example, after the drone lands. The drone may have a Global Positioning System (GPS) and this enables the location of the acquired image to be determined. For example, the orientation of the camera 110 and the position of the drone at the time the image was obtained may be used to determine the geographic footprint of the image on the ground plane. The drone may also have an inertial navigation system, for example based on a laser gyroscope. In addition to being used to determine the orientation of the drone and thus the orientation of the camera, thereby facilitating the determination of when an image is obtained on the ground, the inertial navigation system may work alone without a GPS system to determine the location of the drone by determining motion away from one or more known locations.

The input unit 20 of the device 10 passes the obtained image to the processing unit 30. Image analysis software operates on the processor 30. The image analysis software may use feature extraction, such as edge detection, and object detection analysis, such as identifiable structures (such as railroad tracks, sleepers, trees, level crossings, station platforms). Thus, based on the known locations of objects, e.g., the locations of buildings within the environment, and based on known structural information, e.g., the distance between crossties and the distance between railroad tracks, the processing unit may patch the acquired images to actually create a composite depiction of the environment, which may in fact be superimposed on a geographic map of the environment. Thus, the geographic location of each image may be determined and no information related to the acquired images based on relevant GPS and/or inertial navigation is necessary. However, if there is GPS and/or inertial navigation information available, such an image analysis that can place a particular image at a particular geographic location based solely on the image is not required. However, if GPS and/or inertial navigation based information is available, such image analysis may be used to augment the geographic location associated with the image. Thus, for example, if, on the basis of GPS and/or inertial navigation based information, the center of the acquired image is deemed to be located 22cm from the side edge and 67cm from the end of a particular railroad tie for a section of railroad, whereas from the actual acquired image, the center of the image is determined to be located 25cm from the edge and 64cm from the end of the tie using the aforementioned image analysis, then the location derived based on GPS and/or inertial navigation may be increased by being required to move the location 3cm in one direction and 3cm in the other.

The processor 30 runs further image processing software. This software analyzes the image to determine the areas in the image where vegetation will be found. Vegetation may be detected based on the shape of features within the acquired image, where, for example, boundary detection software is used to delineate the outer boundaries of the object and the outer boundaries of features within the outer boundaries of the object itself. A vegetation image database can be used to help determine whether features in the image are relevant to vegetation, for example using a trained machine learning algorithm, such as an artificial neural network or decision tree analysis. The camera may obtain a multispectral image that has information related to the colors in the image, and this may be used alone or in combination with feature detection to determine where vegetation will be found in the image. As described above, because the geographic location of the image can be determined, one or more locations in the image where vegetation will be found can be mapped to the precise location of the vegetation on the ground, based on knowledge of the size of the image on the ground.

The processor 30 then runs additional image processing software which may be part of the vegetation location determined based on feature extraction (if used). This software includes a machine learning analyzer. Images of specific weeds were obtained, along with information also relating to the size of the weeds being used. Information relating to the geographic location in the world where such weeds are found, and information relating to the time of year when such weeds are found, including when they bloom, etc., may be tagged with the image. The name of the weed was also labeled with an image of the weed. A machine learning analyzer, which may be based on an artificial neural network or a decision tree analyzer, may now be trained on this image where ground truth has been obtained. In this way, when a new image of vegetation is provided to the analyzer, where this image may have an associated time stamp (e.g., time of year) and geographic location (such as germany or south africa) affixed to it, the analyzer determines the particular type of weeds in the image by comparing the image of the weeds found in the new image to the images of the different weeds for which it was trained, where the size of the weeds, and where and when they were growing, are also taken into account. The specific location of that weed type on the ground within the environment, as well as its size, can thus be determined.

The processor 30 has a database containing different weed types, as well as the optimal weed control technique to be used in controlling that weed type. The size of the weeds or weed clumps on the ground is also considered in determining which weed control technique (also called vegetation control technique) to use. For example, chemical spraying can be an optimal weed control technique for a particular type of weed. The processor can then determine that for a single weed or a small clump of such weed chemical spray weed control technology at a particular location in the environment, the weed should be activated at that particular location to control the weeds. However, if it has been determined that there is a large clump of this particular type of weed in the environment, then the processor may determine that to mitigate the effect of the chemical on the environment, a different weed control technique should be used to control a larger clump of that particular weed at a particular location in the environment, such as flame-based weed control, or high voltage-based weed control, or steam or high pressure water-based weed control techniques, or microwave-based weed control techniques. The processor ensures that all weeds that need to be controlled are assigned at least one weed control technique to be used on them. It may be that in order to optimally control a particular type of weed, two different types of weed control techniques, such as microwave radiation and high voltage, should be applied and the processor build a suitable weed control map.

Thus, the drone's camera 110 obtains an image of the environment, which is passed to the processor 30, which determines what weed control technology should be applied at which particular geographic location in the environment. Thus, in effect, a weed map or weed control technique map may be generated that indicates where in the environment a particular weed control technique should be used.

With continued reference to fig. 4, the weed control train 130 is advanced along the railroad track. The weed control train has several truck carriages each accommodating weed control technology. The first boxcar has a chemical spray based weed control technology 120 a. The second boxcar has high voltage-based weed control technology 120b, other weed control technologies are laser-based 120c, microwave-based 120d, steam-based 120e, and other weed control technologies are available, such as flame-based, solid (foam) chemical deposition, and even mechanical-based weed control technologies. The weed control train has a processor (not shown) that uses the weed map or weed control map described above. The weed control train has means to determine its geographical location, which may be based on one or more of GPS, inertial navigation, image analysis to locate the location of the weed control train and the specific location of the weed control technology. This means that different weed control technologies can be activated at specific sites of weeds as the weed control train passes through the environment, where the specific weed control technology activated at the site of the weeds has been determined to be optimal for that task. As discussed above, the weed control train may have a camera and obtain images. The acquired images can be processed by a processor on the weed control train to determine the location of the train itself by determining the location of the features in the crossties and the surrounding environment. In addition, when the weed control train has a GPS and/or inertial system, the GPS and/or inertial navigation system can be used to determine the location of the train so that the correct weed control technology can be activated at the location of a particular weed. However, if the train also has a camera that obtains an image of the surroundings, feature extraction such as the position of the crossties can be used to augment the position determined by GPS and/or inertial navigation to correct the position so that weed control techniques can be activated at the precise location of the weeds to take into account the position obtained from the GPS system, for example. Thus, the image processing required to determine the position of the crosstie can be run quickly while location updates are applied quickly because the image processing complexity in locating features such as railroad crossties is relatively small.

Fig. 5 shows another example of a system 100 for weed control. The system for weed control of fig. 5 is similar to the one shown in fig. 4. However, in FIG. 5 the weed control train 130 has the camera 110 and device 10 previously discussed. The camera 110 on the weed control train 130 now obtains the image previously obtained by the drone. The processor 30 of the plant on the weed control train 130 processes the acquired images to determine the location and type of weeds. It is not required at this time to determine the precise geographical location of the weeds. Rather, based on the relative spacing between the camera 110 and the weed control technology 120 on the train, the obtained image can be located at a specific point on the ground where the weed located and identified in the image is thus also located on the ground, and the desired weed control technology 120a, 120b, 120c, 120d, or 120e will be activated at the determined weed location. Then, from the knowledge of the forward movement of the weed control train (its speed) and the time at which the image was obtained, it can be determined when the required weed control technique should be activated such that it is activated at the location of the weeds. In this way, the weed control train does not need to have GPS and/or inertial navigation systems or image-based absolute geographical location determination means. Rather, to account for the processing required to determine the type of weed and its precise location in the image and its precise location on the ground-within the train coordinate system-the camera 110 must be spaced from the weed control technique 120 by a distance at least equal to the processing time multiplied by the maximum speed of the weed control train during weed control. Thus, for example, if 0.2s, 0.4s or 0.8s is required for train handling traveling at 25m/s, then the camera 110 must be spaced 5m, 10m or 20m in front of the weed control technology 120e for this train speed with reference to FIG. 5. The reduction in train speed makes it possible to reduce the spacing. Also, the camera 110 that obtains the image may have a very short exposure time, thereby minimizing image shift due to train movement during the exposure time. This can be by a variety of different means including, for example, using a camera with short exposure times or short pulse illumination, for example by a laser or LED, in conjunction with a filter. However, the device may use a GPS system and/or an inertial navigation system and/or image analysis to determine the precise geographical location of the weeds. This means that it is possible to determine what weeds have been controlled by what weed control means, and a record of where those weeds are located. Moreover, by generating precise geographical locations of weeds, the weed control technologies 120a-120a can have associated location determination means, such as GPS systems and/or inertial navigation systems and/or image-based systems, that can be used to provide precise locations for particular weed control technologies. Thus, the front cars of the train may have an image acquisition and analysis unit that enables the construction of a weed control map. The last wagons of the train may now have the weed control technology contained within them, with these latter wagons being spaced from the former by tens or hundreds of meters through the load-carrying wagons. The absolute spacing of the front cars to the rear cars may vary as the train goes up and down hills, but because the boxcars with weed control technology know their precise location, when they move forward to the location of weeds or the area of a particular type of weed, the appropriate weed control technology can be activated at that precise geographic location.

Fig. 5 shows two views of the weed control train 130, the upper being a side view and the lower being a plan view. This shows the image-capturing cameras 110 extending between and to either side of the rails. Each freight car of a weed control train has associated weed control technology that can be applied under the train and on both sides of the train.

Fig. 6 shows a boxcar of the weed control train 130 shown in fig. 4-5 having a chemical spray based weed control technique 120 a. Fig. 6 shows a rear view of this boxcar of the train, which is a view along the railway track. Several independent spray nozzles of the weed control technology 120a extend laterally under the train and to both sides of the train. The spray nozzle itself may now have specific controls other than on or off, or may be directionally controlled to spray left and right or downward, and/or controlled such that the angular range of the spray is altered so that, for example, a narrow spray jet may be directed towards a single weed. When one of the spray nozzles passes over a weed and the weed has been identified as a weed that should be controlled by chemical spraying, the processor 30 activates the particular nozzle that sprays the chemical at the particular location of the weed where it is desired to control that weed by chemical spraying. In fig. 6, this weed has two specific locations, one found between the tracks and one on the left side of the track, and thus two spray nozzles are activated. Note that weeds may pass under this boxcar, which has caused one of the other weed control technologies 120b-e to be applied to the weeds.

Fig. 7 shows a freight car of a weed control train 130 as shown in fig. 4-5 having a high voltage based weed control technology 120 b. Fig. 7 shows a rear view of this boxcar of the train, which is a view along the railway track. Pairs of independent electrode pairs of the weed control technology 120b extend laterally under and to both sides of the train, as shown in more detail in fig. 9. When one of the pairs of electrodes passes over a weed that has been identified as a weed that should be controlled by high voltage-based weed control, the processor 30 activates that particular pair of electrode pairs at the particular location of the weed that requires control by high voltage. In fig. 7, there are two specific sites of such weeds, one is a large clump found between the tracks that also extends to the right hand side of the tracks, and one is a small clump found on the left side of the tracks, and thus one electrode pair has been activated on the left hand side and several electrode pairs underneath the train and extending to the right hand side are activated. Note that weeds may have one of the other weed control technologies 120c-e applied to them as they may pass under this truck bed, and weeds may pass under the truck bed in an untreated state if it has been determined that the weeds should be addressed by the chemical spray-based weed control technology 120 a.

Fig. 8 shows a boxcar of the weed control train 130 as shown in fig. 4-5, having a laser-based weed control technique 120 c. Fig. 8 shows a rear view of this boxcar of the train, which is a view along the railway track. Several independent laser systems of the weed control technology 120c extend laterally under the train and to both sides of the train. Each laser system may simply be operated in an on/off state to illuminate an area beneath the laser system or may also be directionally steered as required to illuminate only a specific spot of weed. When one of these laser systems passes over weeds that have been identified as weeds that should be controlled by laser radiation-based weed control, the processor 30 activates the particular laser system at the particular location of the weed that requires control by laser radiation. In fig. 8 there is only one specific spot of such a weed, which is located between the tracks in the vicinity of the left-hand side track, and thus under the train a laser system is activated, which is directed towards the specific spot of the weed. The laser system may be laser diode based, NG: YAG, excimer-based or any other laser system indicated as suitable for weed control. Note that the weeds passing under this truck bed may have one of the other weed control technologies 120d-e applied to them, and if it has been determined that the weeds should be treated by the chemical spray-based weed control technology 120a and/or the high voltage-based weed control technology 120b, the weeds may pass under the truck bed in an untreated state.

Fig. 9 shows more details of the high voltage-based weed control technique 120 b. Pairs of electrodes are provided which, when activated, cause an electrical current to flow from one electrode to the other electrode through the weeds and the ground including the roots of the weeds. One subunit shown in fig. 120b may have one electrode pair or indeed several electrode pairs to enable greater resolution and smaller spatial extent of application of such high voltage-based weed control. The high voltage may be applied for a period of time in a DC mode or for a period of time in an AC mode.

FIG. 10 shows a depiction of a railway environment showing a railway track and the ground on both sides of the track. Several weed fields are shown, with a large clump of weeds of one type having a clump of different types of weeds within the clump. Shown in fig. 10 is a particular weed control technique that has been determined to be activated for these particular weeds. This can be considered as the weed control map discussed with reference to fig. 4, or to determine in real time what weed control techniques should be used where as discussed with reference to fig. 5.

The examples detailed above are discussed with reference to railways, where different weed control technologies (vegetation control technologies) are accommodated in different freight cars of a train. These can be housed within a single truck bed and there can be only two, three or four weed control technologies, such as only chemical spraying and high voltage technologies. In addition, in addition to the weed control train, a truck or van or Unimog (Unimog) may have several weed control technologies installed thereon/therein, and travel around an industrial area or even an area such as an airport based on a previously obtained and processed image or based on an image obtained and processed by itself and apply a specific weed control technology to a specific weed type as described above.

In another exemplary embodiment, a computer program or a computer program element is provided, characterized by being configured to perform the method steps of the method according to one of the preceding embodiments on a suitable system. The computer program element may thus be stored on a computer unit, which may also be part of an embodiment. This computing unit may be configured to perform or cause to be performed the steps of the method described above. Moreover, it may be configured to operate the components of the devices and/or systems described above. The computing unit may be configured to operate automatically and/or to execute user commands. The computer program may be loaded into the working memory of a data processor. The data processor may thus be equipped to carry out a method according to one of the preceding embodiments.

This exemplary embodiment of the invention covers both a computer program that uses the invention from the beginning and a computer program that changes an existing program into a program that uses the invention by means of an update.

Further, the computer program element may be capable of providing all the necessary steps of a process to fulfill the exemplary embodiment of the method described above.

According to another exemplary embodiment of the present invention, a computer-readable medium, such as a CD-ROM, a USB stick or the like, is provided, wherein the computer-readable medium has stored thereon a computer program element, the computer program element being as described in the previous section.

A computer program may be stored and/or distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the internet or wired or wireless telecommunication systems.

However, the computer program may also be provided over a network, such as the world wide web, and may be downloaded into the working memory of a data processor from this network. According to another exemplary embodiment of the present invention, a medium is provided for making available for downloading a computer program element arranged to perform a method according to one of the above described embodiments of the present invention.

It has to be noted that embodiments of the invention are described with reference to different subject matters. In particular, some embodiments are described with reference to method type claims whereas other embodiments are described with reference to apparatus type claims. However, a person skilled in the art will gather from the above and the following description that, unless other notified, also any combination of features belonging to one type of subject matter is possible, apart from any combination between features relating to different subject matters, which combinations are considered to be disclosed with this application. However, all of the features may be combined to provide more synergistic effects than a simple addition of the features.

While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive. The invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims shall not be construed as limiting the scope.