Farmland laser operation planning method and device based on artificial intelligence and electronic equipment

Abstract

The invention provides a farmland laser operation planning method and device based on artificial intelligence and electronic equipment, and relates to the technical field of laser intelligent equipment. According to the specific implementation scheme, the method comprises the steps of obtaining visual related data of a target area and agricultural machine poses; based on the visual related data and the agricultural machine pose, performing biological target detection to obtain a biological target detection result; working path planning is conducted based on the biological target detection result, laser control parameters in the laser working process are determined, and the laser control parameters are used for controlling a laser device to output laser.

Description

Technical Field

[0001] This application relates to the field of laser intelligent equipment technology, and more specifically, to an artificial intelligence-based method, apparatus, and electronic device for planning laser operations in farmland. Background Technology

[0002] Weeds and pests in farmland are core factors that restrict the quality and yield of agricultural production. How to effectively carry out weeding and pest control has become an important technical issue.

[0003] Traditional methods rely heavily on chemical agents, such as herbicides and pesticides, which may lead to environmental pollution and ecological damage, and also result in insufficient operational precision.

[0004] In addition, in traditional methods, weeding and pest control are usually carried out separately, which results in low work efficiency. Summary of the Invention

[0005] This application provides an artificial intelligence-based method, apparatus, and electronic device for farmland laser operation planning, aiming to at least address one of the aforementioned technical deficiencies. The technical solution adopted in this application is as follows:

[0006] In a first aspect, embodiments of this application provide an artificial intelligence-based method for planning laser-based farmland operations, the method comprising:

[0007] Acquire visual data related to the target area and the agricultural machinery's pose;

[0008] Based on the visual relevance data and the agricultural machinery pose, biological target detection is performed to obtain biological target detection results;

[0009] Based on the biological target detection results, the operation path is planned, and the laser control parameters during the laser operation are determined. The laser control parameters are used to control the laser output.

[0010] As an optional approach, the biological target detection based on the visual relevance data and the agricultural machinery pose, to obtain the biological target detection result, includes:

[0011] A biological target detection model is obtained, which includes a feature fusion layer and a biological target detection layer, wherein the biological target detection layer includes a weed detection head, a pest detection head, and a crop detection head;

[0012] The visually relevant data of the target area and the pose of the agricultural machinery are input into the feature fusion layer to obtain the fused features;

[0013] The fused features are input into the biological target detection layer to obtain weed detection results, pest detection results, and crop detection results.

[0014] The step of planning the operation path based on the biological target detection results and determining the laser control parameters during the laser operation includes:

[0015] Based on the weed detection results, pest detection results, and crop detection results, an operation path and laser control parameters corresponding to each sub-path segment in the operation path are generated.

[0016] As an optional approach, inputting the fused features into the biological target detection layer includes:

[0017] The fused features are subjected to feature layering processing to obtain a first feature and a second feature, wherein the first feature is at a higher level than the second feature.

[0018] After performing deconvolution on the first feature, it is fused with the second feature to obtain a cross-layer fused feature;

[0019] The cross-layer fusion features are subjected to dilated convolution processing to obtain enhanced features;

[0020] The enhanced features are input into the pest detection head.

[0021] As an optional approach, the visually relevant data includes the RGB image and depth image of the target region. The step of inputting the visually relevant data of the target region and the agricultural machinery pose into the feature fusion layer to obtain the fused features includes:

[0022] Cross-attention processing is performed on the RGB image and the depth image to obtain visual fusion features;

[0023] The agricultural machinery pose is encoded into agricultural machinery pose features, and the agricultural machinery pose features are concatenated with the visual fusion features to obtain the fused features.

[0024] As an optional approach, the sub-path includes a weeding sub-path, an insect control sub-path, and a cooling sub-path. The weeding sub-path corresponds to weeding control parameters, which are used to control the laser output for weeding. The insect control sub-path corresponds to insect control parameters, which are used to control the laser output for insect control. The cooling sub-path corresponds to cooling control parameters, which are used to control the laser to be in a cooling state.

[0025] As an alternative approach, the above methods also include:

[0026] Based on the results of the operation path planning, the movement of the agricultural machinery is controlled, and during the movement of the agricultural machinery, the laser output is controlled based on the laser control parameters.

[0027] As an alternative approach, the above methods also include:

[0028] The amount of fertilizer to be applied is determined based on the crop test results.

[0029] During the movement of the agricultural machinery, fertilization is carried out based on the amount of fertilizer applied.

[0030] As an optional method, determining the fertilizer application rate based on the crop detection results includes:

[0031] Obtain at least one of the following: soil fertility data, crop growth cycle information, and meteorological data;

[0032] The amount of fertilizer is determined based on at least one of the soil fertility data, the crop growth cycle information, and the meteorological data, and based on the crop detection results.

[0033] Secondly, embodiments of this application provide an artificial intelligence-based farmland laser operation planning device, the device comprising:

[0034] The data acquisition module is used to acquire visually relevant data of the target area and the pose of the agricultural machinery;

[0035] The biological target detection module is used to perform biological target detection based on the visual related data and the agricultural machinery pose, and obtain the biological target detection result;

[0036] The laser operation planning module is used to plan the operation path based on the biological target detection results and determine the laser control parameters during the laser operation process. The laser control parameters are used to control the laser output.

[0037] Thirdly, embodiments of this application provide an electronic device, which includes:

[0038] One or more processors; and

[0039] A memory associated with one or more of the aforementioned processors, the memory being used to store program instructions that, when read and executed by the one or more of the aforementioned processors, perform the steps of the aforementioned artificial intelligence-based farmland laser operation planning method.

[0040] The beneficial effects of the technical solutions provided in this application are:

[0041] The solution provided in this application performs biological target detection based on visual correlation data of the target area and the agricultural machinery pose, obtaining biological target detection results. Based on these results, a work path is planned, and laser control parameters are determined during the laser operation. These laser control parameters are used to control the laser output. This solution integrates multi-source data for accurate biological target detection, improving the accuracy of work path planning and laser control parameters. This enables high-precision laser weeding or laser pest control, avoiding the environmental pollution and ecological damage caused by traditional chemical methods, and contributing to green agriculture. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below.

[0043] Figure 1 A flowchart illustrating the artificial intelligence-based farmland laser operation planning method provided in this application embodiment;

[0044] Figure 2 A schematic diagram of the structure of an artificial intelligence-based farmland laser operation planning device provided in an embodiment of this application;

[0045] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0046] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0047] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0048] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0049] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0050] Figure 1 This illustration shows a flowchart of an artificial intelligence-based laser operation planning method for farmland, as provided in an embodiment of this application. Figure 1 As shown, this method mainly includes:

[0051] Step S110: Obtain visually relevant data of the target area and the pose of the agricultural machinery;

[0052] Step S120: Based on visually relevant data and the agricultural machinery pose, perform biological target detection to obtain biological target detection results;

[0053] Step S130: Based on the biological target detection results, plan the operation path and determine the laser control parameters during the laser operation. The laser control parameters are used to control the laser output.

[0054] The target area can be farmland or other areas that require pest control or weeding.

[0055] Agricultural machinery attitude is used to describe the position and orientation of agricultural machinery in three-dimensional space, including position parameters (such as three-dimensional coordinates) and attitude parameters (such as pitch angle, roll angle and yaw angle).

[0056] Visual data can include multimodal data such as RGB images, depth images, and multispectral data.

[0057] By integrating visually relevant data and agricultural machinery pose, multimodal data can be used for biological target detection, thereby improving the accuracy of biological target detection results.

[0058] Operation path planning can plan the travel path of agricultural machinery in the target area. During the movement of agricultural machinery, the laser outputs laser based on laser control parameters to achieve pest control or weeding operations.

[0059] By planning the operation path based on the results of biological target detection and controlling the laser output through laser control parameters, laser weeding or laser pest control can be achieved. This can achieve high-precision weeding or pest control, avoiding the environmental pollution and ecological damage caused by the use of chemical agents in traditional methods, and contributing to the realization of green agriculture.

[0060] The method provided in this application performs biological target detection based on visual correlation data of the target area and the agricultural machinery pose, obtains biological target detection results, plans the operation path based on the biological target detection results, and determines the laser control parameters during the laser operation. These laser control parameters are used to control the laser output. Based on this solution, accurate biological target detection can be achieved by fusing multi-source data, improving the accuracy of operation path planning and laser control parameters, thereby realizing high-precision laser weeding or laser pest control. This avoids the environmental pollution and ecological damage caused by the use of chemical agents in traditional methods, contributing to the realization of green agriculture.

[0061] In one optional embodiment of this application, biological target detection is performed based on visually relevant data and the agricultural machinery pose to obtain biological target detection results, including:

[0062] A biological target detection model is obtained, which includes a feature fusion layer and a biological target detection layer. The biological target detection layer includes a weed detection head, a pest detection head, and a crop detection head.

[0063] The visually relevant data of the target area and the pose of the agricultural machinery are input into the feature fusion layer to obtain the fused features.

[0064] The fused features are input into the biological target detection layer to obtain weed detection results, pest detection results, and crop detection results.

[0065] Based on the results of biological target detection, the operation path is planned, and the laser control parameters during the laser operation are determined, including:

[0066] Based on the results of weed detection, pest detection, and crop detection, an operation path and laser control parameters corresponding to each sub-path segment within the operation path are generated.

[0067] In the biological target detection model, a feature fusion layer serves as the backbone network, extracting fused features from multi-source input data. The biological target detection layer consists of weed detection heads, pest detection heads, and crop detection heads, used to detect the location and category of the corresponding biological targets, respectively.

[0068] Based on the results of weed detection, pest detection, and crop detection, operation path planning is carried out. Sub-paths can be designed for different biological targets, and corresponding laser control parameters can be generated to control agricultural machinery to carry out targeted operations.

[0069] For example, in this solution, the laser can provide at least dual-band laser light for both weeding and pest control. The laser can be positioned with high precision using a galvanometer system.

[0070] In this solution, sub-paths for weeding and pest control are provided in the operation path, and corresponding laser control parameters are designed to control the laser, so that weeding and pest control can be completed simultaneously in one operation, which greatly improves the operation efficiency compared to independent operation.

[0071] In one optional embodiment of this application, inputting the fused features into the biological target detection layer includes:

[0072] The fused features are subjected to feature layering processing to obtain a first feature and a second feature, wherein the first feature is at a higher level than the second feature.

[0073] After performing deconvolution on the first feature, it is fused with the second feature to obtain a cross-layer fused feature;

[0074] The cross-layer fusion features are subjected to dilated convolution processing to obtain enhanced features;

[0075] The enhanced features are input into the pest detection head.

[0076] The fused features are processed using a feature pyramid to obtain multi-scale feature maps. From these maps, a first feature and a second feature are extracted. The first feature is a high-level feature that contains strong semantic information (such as the insect category) but loses detail. The second feature is a low-level feature that retains more morphological details but has more semantic noise.

[0077] Perform deconvolution processing on the first feature (i.e., the high-level feature). For example, use a transposed convolution kernel (such as a 4×4 kernel with a stride of 2) to perform inverse spatial expansion on the first feature to restore high resolution.

[0078] The first and second features obtained after deconvolution are fused, for example, by element-wise addition, to obtain cross-layer fused features. In cross-layer fused features, high-level semantics can guide the localization of low-level details (such as guiding the location of insect antennae), while low-level details correct misclassifications of high-level semantics (such as misclassifying leaf shadows as insect bodies).

[0079] Applying dilated convolution to cross-layer fusion features can expand the receptive field of view and avoid fragmentation of local features (such as detecting only a part of the insect body), so as to cover the overall morphology of the insect body and improve the accuracy of pest localization.

[0080] In one optional embodiment of this application, the visually relevant data includes an RGB image and a depth image of the target region. The visually relevant data of the target region and the agricultural machinery pose are input into a feature fusion layer to obtain fused features, including:

[0081] Cross-attention processing is performed on RGB and depth images to obtain visual fusion features;

[0082] The agricultural machinery pose is encoded into agricultural machinery pose features, and the agricultural machinery pose features are spliced ​​with visual fusion features to obtain fused features.

[0083] The visually relevant data includes RGB and depth images of the target region, and the semantic information of the RGB and depth images can be aligned through a cross-attention mechanism.

[0084] For example, the features of an RGB image can be linearly transformed and used as a Query, and the features of a depth image can be linearly transformed and used as a Key / Value.

[0085] Agricultural machinery pose data (such as GPS coordinates and orientation) is encoded into agricultural machinery pose features, and these features are then concatenated with visual features along the channel dimension.

[0086] In one optional embodiment of this application, the sub-path includes a weeding sub-path, an insect control sub-path, and a cooling sub-path. The weeding sub-path corresponds to weeding control parameters, which are used to control the laser output for weeding. The insect control sub-path corresponds to insect control parameters, which are used to control the laser output for insect control. The cooling sub-path corresponds to cooling control parameters, which are used to control the laser to be in a cooling state.

[0087] The sub-paths in the operation path can specifically include weeding sub-paths, pest control sub-paths, and cooling sub-paths.

[0088] When the agricultural machinery moves along the weeding section path, the laser can output laser light for weeding based on the weeding control parameters, thereby realizing weeding operations along the weeding section path.

[0089] The weeding route can be determined based on the weed detection results. For example, the weed detection results include the location of the weeds, and the weeding route can be planned based on the location of the weeds.

[0090] As the agricultural machinery moves along the insect-removal section path, the laser can output laser light for insect removal based on the insect-removal control parameters, thereby enabling weeding operations to be carried out along the insect-removal section path.

[0091] The pest control route can be determined based on the pest detection results. For example, the pest detection results include the location of the pests, and the pest control route can be planned based on the location of the pests.

[0092] When the agricultural machinery moves along the cooling section path, the laser can be in a cooling state, that is, in a non-working state, so that the laser can be cooled down.

[0093] Cooling segment paths can be determined based on crop detection results. For example, crop detection results include the location of the crops, and cooling segment paths can be planned based on the crop locations.

[0094] In one optional embodiment of this application, the above solution further includes:

[0095] Based on the results of the operation path planning, the movement of agricultural machinery is controlled, and during the movement of agricultural machinery, the laser output is controlled based on the laser control parameters.

[0096] For example, the laser can also output lasers for weeding different types of weeds or for pest control of different types of pests.

[0097] In this example, the weed detection results can include the location and type of weeds, weeding segment paths can be planned based on the location of weeds, and corresponding laser control parameters can be determined based on the type of weeds.

[0098] The pest detection results include the location and type of pests. Weeding routes can be planned based on the location of pests, and corresponding laser control parameters can be determined based on the type of pest.

[0099] In one optional embodiment of this application, the above method further includes:

[0100] Fertilizer application rate is determined based on crop testing results;

[0101] Fertilization operations are carried out based on the amount of fertilizer applied during the movement of agricultural machinery.

[0102] Among them, crop testing results can determine information such as crop type and planting density, thereby enabling the determination of fertilizer application based on crop type and planting density, and thus carrying out fertilization operations based on fertilizer application.

[0103] In one optional embodiment of this application, determining the fertilizer application amount based on crop testing results includes:

[0104] Obtain at least one of the following: soil fertility data, crop growth cycle information, and meteorological data;

[0105] The amount of fertilizer is determined based on at least one of the following: soil fertility data, crop growth cycle information, and meteorological data, and based on crop testing results.

[0106] The amount of fertilizer applied is also related to factors such as soil fertility data, crop growth cycle information, and meteorological data. Therefore, the amount of fertilizer applied can be determined based on at least one of the following: soil fertility data, crop growth cycle information, and meteorological data, and based on crop test results.

[0107] For example, a pre-trained fertilizer application prediction model can be provided, into which soil fertility data, crop growth cycle information, meteorological data and crop detection results are input to obtain the prediction results output by the fertilizer application prediction model.

[0108] The method provided in this application can be applied to an intelligent laser agricultural machinery platform, which can be subordinate to a smart planting system. The smart planting system can also include: a laser plasma nano-fertilizer preparation device, a seed intelligent coating system, etc.

[0109] The laser plasma nano-fertilizer preparation device generates plasma by bombarding a target material with a pulsed laser, and synthesizes slow-release compound fertilizer.

[0110] The intelligent seed coating system integrates machine vision (to identify seeds with a damage rate >5%) with electrostatic atomization spraying technology to achieve precise control of the nano-fertilizer layer thickness within ±5μm.

[0111] Based on and Figure 1 The method shown follows the same principle. Figure 2 This illustration shows a structural schematic diagram of an artificial intelligence-based farmland laser operation planning device according to an embodiment of this application. Figure 2 As shown, the AI-based farmland laser operation planning device 20 may include:

[0112] Data acquisition module 210 is used to acquire visually relevant data of the target area and the pose of the agricultural machinery;

[0113] The biological target detection module 220 is used to perform biological target detection based on visual related data and agricultural machinery pose, and obtain biological target detection results;

[0114] The laser operation planning module 230 is used to plan the operation path based on the biological target detection results and determine the laser control parameters during the laser operation process. The laser control parameters are used to control the laser output.

[0115] The device provided in this application performs biological target detection based on visual correlation data of the target area and the agricultural machinery pose, obtains biological target detection results, plans the operation path based on the biological target detection results, and determines the laser control parameters during the laser operation process. These laser control parameters are used to control the laser output. Based on this solution, accurate biological target detection can be achieved by fusing multi-source data, improving the accuracy of operation path planning and laser control parameters, thereby realizing high-precision laser weeding or laser pest control. This avoids the environmental pollution and ecological damage caused by the use of chemical agents in traditional methods, contributing to the realization of green agriculture.

[0116] As an optional approach, the biological target detection module 220 is specifically used for:

[0117] A biological target detection model is obtained, which includes a feature fusion layer and a biological target detection layer. The biological target detection layer includes a weed detection head, a pest detection head, and a crop detection head.

[0118] The visually relevant data of the target area and the pose of the agricultural machinery are input into the feature fusion layer to obtain the fused features.

[0119] The fused features are input into the biological target detection layer to obtain weed detection results, pest detection results, and crop detection results.

[0120] Based on the results of biological target detection, the operation path is planned, and the laser control parameters during the laser operation are determined, including:

[0121] Based on the results of weed detection, pest detection, and crop detection, an operation path and laser control parameters corresponding to each sub-path segment within the operation path are generated.

[0122] As an optional approach, when the biometric target detection module 220 inputs the fused features into the biometric target detection layer, it is specifically used for:

[0123] The fused features are subjected to feature hierarchical processing to obtain the first feature and the second feature, with the first feature having a higher hierarchical level than the second feature.

[0124] After performing deconvolution on the first feature, it is fused with the second feature to obtain a cross-layer fused feature.

[0125] The cross-layer fusion features are processed by dilated convolution to obtain the enhanced features;

[0126] The enhanced features are input into the pest detection head.

[0127] As an optional approach, the visually relevant data includes the RGB image and depth image of the target region. When the biological target detection module 220 inputs the visually relevant data of the target region and the agricultural machinery pose into the feature fusion layer to obtain the fused features, it is specifically used for:

[0128] Cross-attention processing is performed on RGB and depth images to obtain visual fusion features;

[0129] The agricultural machinery pose is encoded into agricultural machinery pose features, and the agricultural machinery pose features are spliced ​​with visual fusion features to obtain fused features.

[0130] As an optional approach, the sub-path includes a weeding sub-path, an insect control sub-path, and a cooling sub-path. The weeding sub-path corresponds to weeding control parameters, which are used to control the laser output for weeding. The insect control sub-path corresponds to insect control parameters, which are used to control the laser output for insect control. The cooling sub-path corresponds to cooling control parameters, which are used to control the laser to be in a cooling state.

[0131] As an alternative, the above-mentioned device further includes:

[0132] The operation module (not shown in the figure) controls the movement of agricultural machinery based on the results of the operation path planning, and controls the laser to output laser based on the laser control parameters during the movement of agricultural machinery.

[0133] As an optional embodiment, the above-mentioned device also includes a fertilizer application module (not shown in the figure), specifically used for:

[0134] Fertilizer application rate is determined based on crop testing results;

[0135] Fertilization operations are carried out based on the amount of fertilizer applied during the movement of agricultural machinery.

[0136] As an optional approach, the fertilization treatment module (not shown in the figure) is specifically used to determine the fertilization amount based on crop detection results for:

[0137] Obtain at least one of the following: soil fertility data, crop growth cycle information, and meteorological data;

[0138] The amount of fertilizer is determined based on at least one of the following: soil fertility data, crop growth cycle information, and meteorological data, and based on crop testing results.

[0139] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0140] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0141] This application also provides an electronic device, including:

[0142] One or more processors; and

[0143] A memory associated with one or more of the aforementioned processors, the memory being used to store program instructions that, when read and executed by one or more of the aforementioned processors, perform the steps of the method in any of the foregoing method embodiments.

[0144] The electronic device provided in this application, compared with the prior art, performs biological target detection based on visual correlation data of the target area and the agricultural machinery pose, obtains biological target detection results, plans the operation path based on the biological target detection results, and determines the laser control parameters during the laser operation process. These laser control parameters are used to control the laser output. Based on this solution, multi-source data can be fused for accurate biological target detection, improving the accuracy of operation path planning and laser control parameters, thereby achieving high-precision laser weeding or laser pest control. This avoids the environmental pollution and ecological damage caused by the use of chemical agents in traditional methods, contributing to the realization of green agriculture.

[0145] As an example, Figure 3 A schematic diagram of the structure of an electronic device to which this application embodiment applies is shown, such as... Figure 3 As shown, the electronic device 2000 includes a processor 2001 and a memory 2003. The processor 2001 and the memory 2003 are connected, for example, via a bus 2002. Optionally, the electronic device 2000 may also include a transceiver 2004. It should be noted that in practical applications, the transceiver 2004 is not limited to one type, and the structure of the electronic device 2000 does not constitute a limitation on the embodiments of this application.

[0146] In this embodiment, the processor 2001 is used to implement the method shown in the above method embodiment. The transceiver 2004 may include a receiver and a transmitter. In this embodiment, the transceiver 2004 is used to enable the electronic device of this embodiment to communicate with other devices during execution.

[0147] Processor 2001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), a FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 2001 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0148] Bus 2002 may include a pathway for transmitting information between the aforementioned components. Bus 2002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 2002 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0149] The memory 2003 may be ROM (Read-Only Memory) or other types of static storage devices capable of storing static information and instructions, RAM (Random Access Memory) or other types of dynamic storage devices capable of storing information and instructions, or EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed discs, laser discs, optical discs, digital universal discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0150] Optionally, the memory 2003 is used to store application code that executes the scheme of this application, and its execution is controlled by the processor 2001. The processor 2001 is used to execute the application code stored in the memory 2003 to implement the steps of the method in any of the foregoing method embodiments.

[0151] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0152] The above description is only a partial embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for planning laser-based farmland operations based on artificial intelligence, characterized in that, include: Acquire visual data related to the target area and the agricultural machinery's pose; Based on the visual relevance data and the agricultural machinery pose, biological target detection is performed to obtain biological target detection results; Based on the biological target detection results, the operation path is planned, and the laser control parameters during the laser operation are determined. The laser control parameters are used to control the laser output.

2. The method according to claim 1, characterized in that, The process of performing biological target detection based on the visual relevance data and the agricultural machinery pose to obtain biological target detection results includes: A biological target detection model is obtained, which includes a feature fusion layer and a biological target detection layer, wherein the biological target detection layer includes a weed detection head, a pest detection head, and a crop detection head; The visually relevant data of the target area and the pose of the agricultural machinery are input into the feature fusion layer to obtain the fused features; The fused features are input into the biological target detection layer to obtain weed detection results, pest detection results, and crop detection results. The step of planning the operation path based on the biological target detection results and determining the laser control parameters during the laser operation includes: Based on the weed detection results, pest detection results, and crop detection results, an operation path and laser control parameters corresponding to each sub-path segment in the operation path are generated.

3. The method according to claim 2, characterized in that, The step of inputting the fused features into the biological target detection layer includes: The fused features are subjected to feature layering processing to obtain a first feature and a second feature, wherein the first feature is at a higher level than the second feature. After performing deconvolution on the first feature, it is fused with the second feature to obtain a cross-layer fused feature; The cross-layer fusion features are subjected to dilated convolution processing to obtain enhanced features; The enhanced features are input into the pest detection head.

4. The method according to claim 2, characterized in that, The visually relevant data includes the RGB image and depth image of the target region. The step of inputting the visually relevant data of the target region and the agricultural machinery pose into the feature fusion layer to obtain fused features includes: Cross-attention processing is performed on the RGB image and the depth image to obtain visual fusion features; The agricultural machinery pose is encoded into agricultural machinery pose features, and the agricultural machinery pose features are concatenated with the visual fusion features to obtain the fused features.

5. The method according to claim 2, characterized in that, The sub-path includes a weeding sub-path, an insect control sub-path, and a cooling sub-path. The weeding sub-path corresponds to weeding control parameters, which are used to control the laser to output laser light for weeding. The insect control sub-path corresponds to insect control parameters, which are used to control the laser to output laser light for insect control. The cooling sub-path corresponds to cooling control parameters, which are used to control the laser to be in a cooling state.

6. The method according to any one of claims 2-5, characterized in that, Also includes: Based on the results of the operation path planning, the movement of the agricultural machinery is controlled, and during the movement of the agricultural machinery, the laser output is controlled based on the laser control parameters.

7. The method according to claim 6, characterized in that, Also includes: The amount of fertilizer to be applied is determined based on the crop test results. During the movement of the agricultural machinery, fertilization is carried out based on the amount of fertilizer applied.

8. The method according to claim 7, characterized in that, Determining the fertilizer application rate based on the crop detection results includes: Obtain at least one of the following: soil fertility data, crop growth cycle information, and meteorological data; The amount of fertilizer is determined based on at least one of the soil fertility data, the crop growth cycle information, and the meteorological data, and based on the crop detection results.

9. A farmland laser operation planning device based on artificial intelligence, characterized in that, include: The data acquisition module is used to acquire visually relevant data of the target area and the pose of the agricultural machinery; The biological target detection module is used to perform biological target detection based on the visual related data and the agricultural machinery pose, and obtain the biological target detection result; The laser operation planning module is used to plan the operation path based on the biological target detection results and determine the laser control parameters during the laser operation process. The laser control parameters are used to control the laser output.

10. An electronic device, characterized in that, include: One or more processors; as well as A memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method according to any one of claims 1-8.

Images