Corn emergence rate and emergence uniformity evaluation method and device, terminal and medium

Abstract

The invention discloses a corn emergence rate and emergence uniformity evaluation method and device, a terminal and a medium. The method comprises the following steps: determining a plurality of corn plant images according to a remote sensing image of a target area; using an improved YOLOv10 model to predict a rotation bounding box and a key point corresponding to each corn plant image, wherein the key point is the center of the corn plant; and determining the corn emergence rate and the emergence uniformity corresponding to the target area according to the rotation bounding box and the key point corresponding to each corn plant image. According to the method, the rotation bounding box and the key points of the corn plants are predicted through the improved YOLOv10 model, the rotation bounding box can effectively reduce interference of a background or other non-crop areas, the detection precision is improved, the centers of the corn plants are accurately marked through the key points, the distance and the distribution condition between the plants can be accurately evaluated, and the detection accuracy is improved. The problem that in the prior art, the center of the corn plant is marked through the detected center point of the target frame, and consequently the detection error is large is solved.

Description

Technical Field

[0001] The invention relates to the field of remote sensing, and in particular to a method, a device, a terminal and a medium for evaluating corn emergence rate and emergence uniformity. Background Art

[0002] Corn is one of the major food crops widely grown around the world, and its production is related to food security and economic benefits. Evaluating corn emergence rate and uniformity can help farmers and agricultural technicians quickly identify problems in the planting process, such as improper sowing depth, poor soil conditions, and low seed quality, so that they can take remedial measures to ensure higher crop yields and more rational resource utilization.

[0003] By using drones to carry cameras to take high-resolution images of corn from the air, agricultural researchers can quickly obtain information about corn in large areas of farmland. At present, image recognition algorithms are usually used to process drone remote sensing data. For example, target detection methods are used to automatically label corn, frame each corn plant in a target frame, and mark the center of the corn by the center point of the detected target frame. However, the actual center of the corn is often not at the geometric center of the target frame, especially when the seedlings are uneven or the plants are blocked, resulting in large detection errors.

[0004] Therefore, the existing technology still needs to be improved and developed. Summary of the invention

[0005] The technical problem to be solved by the present invention is that, in view of the above-mentioned defects of the prior art, a method, device, terminal and medium for evaluating corn emergence rate and uniformity are provided, aiming to solve the problem that the prior art marks the center of the corn plant by detecting the center point of the target frame, resulting in large detection errors.

[0006] The technical solution adopted by the present invention to solve the problem is as follows:

[0007] In a first aspect, an embodiment of the present invention provides a method for evaluating corn emergence rate and emergence uniformity, wherein the method comprises:

[0008] Acquire a remote sensing image of a target area, and determine a plurality of corn plant images based on the remote sensing image;

[0009] Using an improved YOLOv10 model to make predictions based on the plurality of corn plant images, determining a rotation bounding box and key points corresponding to each of the corn plant images, wherein the improved YOLOv10 model is implemented by introducing a Mamba attention module in a neck network layer of the YOLOv10 model, and the key points are used to mark the center of the corn plant;

[0010] The corn germination rate and germination uniformity corresponding to the target area are determined according to the rotation bounding box and key points corresponding to each of the corn plant images.

[0011] In one implementation method, determining a plurality of corn plant images based on the remote sensing image includes:

[0012] Extracting straight lines between rows of corn plants in the remote sensing image, and determining rows of corn plants in the remote sensing image based on the straight lines between rows;

[0013] The remote sensing image is segmented according to the corn plant rows to determine a plurality of corn plant images.

[0014] In one implementation method, extracting straight lines between rows of corn plants in the remote sensing image includes:

[0015] Performing image enhancement on the corn plants in the remote sensing image, and converting the enhanced remote sensing image into a binary image;

[0016] Morphological operations and Hough transforms are performed on the binary image to extract the straight lines between the rows of corn plants in the remote sensing image.

[0017] In one implementation method, the training method of the improved YOLOv10 model includes:

[0018] Acquire a number of historical remote sensing images, annotate each of the historical remote sensing images with a rotating bounding box and key points, and use the annotated historical remote sensing images as training data;

[0019] Constructing a loss function corresponding to the improved YOLOv10 model;

[0020] The improved YOLOv10 model is trained according to the training data and the loss function.

[0021] In one implementation method, constructing the loss function corresponding to the improved YOLOv10 model includes:

[0022] Calculating the classification loss, bounding box loss, and key point loss during the training of the improved YOLOv10 model;

[0023] The classification loss, the bounding box loss, and the key point loss are weighted to determine a loss function corresponding to the improved YOLOv10 model.

[0024] In an implementation method, determining the corn emergence rate corresponding to the target area according to the rotation bounding box corresponding to each corn plant image includes:

[0025] Determining the number of rotation frames corresponding to the target area according to the rotation boundary frames corresponding to each of the corn plant images;

[0026] The area of ​​the target area is obtained, and the corn emergence rate of the target area is determined according to the number of rotating boxes and the area of ​​the target area.

[0027] In one implementation method, determining the uniformity of seedling emergence corresponding to the target area according to the key points corresponding to each of the corn plant images includes:

[0028] Determining the pixel distance between adjacent corn plants according to the key points corresponding to the corn plant images;

[0029] Acquire a ground sampling distance, and determine the actual distance between adjacent corn plants according to each pixel distance and the ground sampling distance;

[0030] The seedling emergence uniformity corresponding to the target area is determined according to the actual distance between each adjacent corn plant.

[0031] In a second aspect, an embodiment of the present invention further provides a device for evaluating corn germination rate and germination uniformity, wherein the device for evaluating corn germination rate and germination uniformity comprises:

[0032] An acquisition module, used for acquiring a remote sensing image of a target area, and determining a plurality of corn plant images according to the remote sensing image;

[0033] A prediction module, configured to use an improved YOLOv10 model to perform prediction based on the plurality of corn plant images, and determine a rotation bounding box and key points corresponding to each of the corn plant images, wherein the improved YOLOv10 model is implemented by introducing a Mamba attention module in a neck network layer of the YOLOv10 model, and the key points are used to mark the center of the corn plant;

[0034] An evaluation module is used to determine the corn seedling emergence rate and seedling emergence uniformity corresponding to the target area according to the rotation boundary box and key points corresponding to each of the corn plant images.

[0035] In a third aspect, an embodiment of the present invention further provides a terminal, comprising a memory and one or more processors; the memory stores one or more programs; the program comprises instructions for executing any of the above-described methods for evaluating corn emergence rate and uniformity; and the processor is used to execute the program.

[0036] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium on which a plurality of instructions are stored, wherein the instructions are suitable for being loaded and executed by a processor to implement any of the above-mentioned methods for evaluating corn emergence rate and emergence uniformity.

[0037] Beneficial effects of the present invention: The embodiment of the present invention determines a number of corn plant images based on the remote sensing image of the target area; uses an improved YOLOv10 model to predict the rotating bounding box and key points corresponding to each corn plant image, and the key point is the center of the corn plant; and determines the corn emergence rate and the emergence uniformity corresponding to the target area according to the rotating bounding box and key points corresponding to each corn plant image. Since the present invention predicts the rotating bounding box and key points of the corn plant through the improved YOLOv10 model, the rotating bounding box can effectively reduce the interference of the background or other non-crop areas, improve the detection accuracy, and the key points can accurately mark the center of the corn plant, and can more accurately evaluate the distance and distribution between plants. Therefore, it can effectively solve the problem that the center of the corn plant is marked by the center point of the detected target box in the prior art, resulting in a large detection error. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 The present invention is a flowchart of a method for evaluating corn emergence rate and emergence uniformity provided in an embodiment of the present invention.

[0040] Figure 2 The present invention provides a flow chart of an embodiment of a method for evaluating corn emergence rate and emergence uniformity.

[0041] Figure 3 It is a diagram of the improved YOLOv10 model architecture provided by an embodiment of the present invention.

[0042] Figure 4 It is a diagram of the Mamba attention module architecture provided by an embodiment of the present invention.

[0043] Figure 5 It is a schematic diagram of the rotation bounding box and key point prediction of a small corn plant provided by an embodiment of the present invention.

[0044] Figure 6 It is a schematic diagram of the rotation bounding box and key point prediction of a large corn plant provided by an embodiment of the present invention.

[0045] Figure 7 It is a schematic diagram of the rotation bounding box and key point prediction of a corn plant containing an interfering plant provided by an embodiment of the present invention.

[0046] Figure 8 It is a schematic diagram of the internal modules of the device for evaluating corn emergence rate and emergence uniformity provided in an embodiment of the present invention.

[0047] Fig. 9 It is a principle block diagram of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The present invention discloses a method, device, terminal and medium for evaluating corn emergence rate and emergence uniformity. In order to make the purpose, technical scheme and effect of the present invention clearer and more specific, the present invention is further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0049] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0050] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as herein.

[0051] Corn is one of the major food crops widely grown around the world, and its production is related to food security and economic benefits. Evaluating corn emergence rate and uniformity can help farmers and agricultural technicians quickly identify problems in the planting process, such as improper sowing depth, poor soil conditions, and low seed quality, so that they can take remedial measures to ensure higher crop yields and more rational resource utilization.

[0052] By using drones to carry cameras to take high-resolution images of corn from the air, agricultural researchers can quickly obtain information about corn in large areas of farmland. At present, image recognition algorithms are usually used to process drone remote sensing data. For example, target detection methods are used to automatically label corn, frame each corn plant in a target frame, and mark the center of the corn by the center point of the detected target frame. However, the actual center of the corn is often not at the geometric center of the target frame, especially when the seedlings are uneven or the plants are blocked, resulting in large detection errors.

[0053] In view of the above-mentioned defects of the prior art, the present invention provides a method for evaluating corn emergence rate and emergence uniformity, the method determining a plurality of corn plant images according to a remote sensing image of a target area; using an improved YOLOv10 model to predict a rotating bounding box and key points corresponding to each corn plant image, the key point being the center of the corn plant; and determining the corn emergence rate and the emergence uniformity corresponding to the target area according to the rotating bounding box and key points corresponding to each corn plant image. Since the present invention predicts the rotating bounding box and key points of the corn plant by using an improved YOLOv10 model, the rotating bounding box can effectively reduce interference from the background or other non-crop areas, improve detection accuracy, and the key points can accurately mark the center of the corn plant, and can more accurately evaluate the distance and distribution between plants, thus effectively solving the problem that the center of the corn plant is marked by the center point of the detected target box in the prior art, resulting in a large detection error.

[0054] Exemplary Methods

[0055] like Figure 1 As shown, the method includes:

[0056] Step S100: acquiring a remote sensing image of a target area, and determining a plurality of corn plant images according to the remote sensing image.

[0057] In short, this embodiment uses a drone equipped with an RGB camera to obtain high-resolution remote sensing images of the target area. Compared with satellite optical and radar remote sensing technologies, the drone equipped with an RGB camera to collect remote sensing images has the characteristics of high mobility and high precision, which is conducive to timely acquisition of information and image interpretation. The drone captures clear image data by flying at low altitudes to ensure that the state of each corn seedling can be effectively recorded, and the use of drones equipped with RGB cameras for image acquisition can cover a large area of ​​farmland in a short time, with high efficiency and good precision, which is suitable for large-scale farmland monitoring. In addition to drones, ground mobile robots or low-altitude aircraft (such as balloons, gliders, etc.) can also be used for image acquisition according to different environmental conditions. The collected image data can select multispectral images, thermal imaging and other image data according to different environmental conditions.

[0058] After acquiring the remote sensing image of the target area, the corn plant image in the remote sensing image is extracted so as to mark the rotation bounding box and key points of each corn plant image.

[0059] In one implementation, determining a plurality of corn plant images according to the remote sensing image includes:

[0060] Step S101, extracting straight lines between rows of corn plants in the remote sensing image, and determining rows of corn plants in the remote sensing image according to the straight lines between rows;

[0061] Step S102: segment the remote sensing image according to the corn plant rows to determine a plurality of corn plant images.

[0062] Specifically, corn is usually planted in rows. After extracting the straight lines between the rows of corn, the planting rows of crops in the target area are quickly detected based on the straight lines between the rows of corn to obtain the rows of corn plants, thereby reducing the interference of irrelevant factors such as soil on subsequent detection.

[0063] According to the identified corn plant rows, the original high-resolution remote sensing image is segmented to ensure that each corn plant in the remote sensing image remains intact in the cropped image, such as Figure 2 shown.

[0064] In order to improve the accuracy of subsequent annotation of rotated bounding boxes and key points of corn plants, this embodiment also performs image enhancement on the segmented corn plant images to improve the quality and diversity of the corn plant images, such as: performing image preprocessing according to different weather and lighting conditions to reduce the impact of the environment on the detection results, and splicing multiple images to increase data diversity; generating new images through linear interpolation to enhance the model's learning of gradient information; and randomly adjusting the brightness, contrast, saturation and hue of the image to improve the model's reasoning ability under different lighting conditions.

[0065] In one implementation, extracting straight lines between rows of corn plants in the remote sensing image includes:

[0066] Step S1011, performing image enhancement on the corn plants in the remote sensing image, and converting the enhanced remote sensing image into a binary image;

[0067] Step S1012: performing morphological operations and Hough transform on the binary image to extract the straight lines between the rows of corn plants in the remote sensing image.

[0068] Specifically, this embodiment enhances the image of corn plants in the remote sensing image through the green channel or the normalized vegetation index to highlight the green area of ​​the corn plants. The Normalized Difference Vegetation Index (NDVI index) is a remote sensing index that reflects the vegetation coverage of the surface. The calculation formula is: NDVI = (NIR-R) / (NIR+R), where NIR is the reflection value of the near infrared band and R is the reflection value of the red light band.

[0069] For the enhanced remote sensing image, the crop is distinguished from the background by using the color or texture features of the vegetation (such as green channel, normalized vegetation index, etc.). This embodiment uses threshold segmentation technology to generate a binary image from the remote sensing image, in which the crop is white and the background is black.

[0070] The binary image is smoothed by morphological operations (such as opening or closing operations) to remove noise and small areas and retain continuous corn plant areas. The binary image after the morphological operation is then subjected to Hough transform to extract the straight lines between the rows of corn plants so as to quickly detect the median rows of corn plants in the target area.

[0071] Step S200, using an improved YOLOv10 model to make predictions based on the plurality of corn plant images, and determining a rotation bounding box and key points corresponding to each of the corn plant images, wherein the improved YOLOv10 model is implemented by introducing a Mamba attention module in a neck network layer of the YOLOv10 model, and the key points are used to mark the center of the corn plant.

[0072] The YOLOv10 model has high-speed and high-precision detection capabilities. The model proposes a lightweight classification head, spatial-channel decoupled downsampling, and sort-guided block design to reduce significant computational redundancy and achieve a more efficient architecture. In addition, to improve accuracy, the model explores large kernel convolution and proposes an effective partial self-attention (PSA) module to enhance the model's understanding ability.

[0073] like Figure 3 As shown, this embodiment introduces the Mamba attention module into the neck network layer of the YOLOv10 model to obtain an improved YOLOv10 model. The Mamba attention module is as follows Figure 4As shown in the figure. The Mamba attention module combines position embedding to promote accurate position-based visual understanding. By combining the Mamba module, the YOLOv10 model can be quickly built and optimized, and the linear complexity of CNN and the global receptive field of the Mamba attention module can be used to effectively detect the emergence of corn plants. It not only improves the detection accuracy of the model in key areas, but also significantly optimizes the running speed of the model and reduces the memory usage during the calculation process. The improved YOLOv10 model can significantly improve the efficiency and resource utilization of the model while maintaining high accuracy, which is particularly suitable for real-time monitoring and evaluation tasks of large-scale farmland.

[0074] Specifically, Mamba is a sequence modeling architecture that leverages a selective state-space model (SSM) to enable efficient processing of long sequences. It addresses the computational inefficiency of the Transformer architecture, especially when processing long sequences, by introducing a mechanism that allows the model to selectively propagate or forget information based on the current input. This is achieved by making the SSM parameters a function of the input, thereby enhancing the model's ability to perform content-based reasoning. Mamba operates in two modes: recurrent mode for autoregressive inference and convolutional mode for parallel training. The architecture is simplified by integrating the selective SSM into a single block, without the need for attention or multi-layer perceptron (MLP) components, resulting in a streamlined design that maintains linear time complexity in the length of the sequence.

[0075] The Mamba attention module combines adaptive average pooling and the Mamba module to enhance the feature representation capability of deep learning models. The module takes an input tensor x∈R B×C×H×W , where B represents the batch size, C represents the number of channels, and H and W represent the height and width respectively. This module generates two feature representations x by adaptive average pooling along the width and height dimensions. x and x y :

[0076] x x =AdaptiveAvgPool2d(x,(None,1)),

[0077] x y =AdaptiveAvgPool2d(x,(1,None)),

[0078] In this process, x x Compress the width of the input feature map while retaining the height information, x y Compress the height while retaining the width information. The module uses Mamba to process the above features and extract important information. x , the output is represented by x ma :

[0079] x ma =Mamba(x x ),

[0080] Apply the Sigmoid activation function to introduce nonlinearity and generate dynamic feature weights:

[0081] x x =σ(x ma ),

[0082] x y =σ(Mamba(x y )),

[0083] x x and x y Represents the weights generated in the width and height dimensions, allowing the model to dynamically adjust the attention to different regions of the input features. Finally, the Mamba attention module combines the input tensor with the processed weights through element-wise multiplication to obtain the output:

[0084]

[0085] where ⊙ represents element-wise multiplication, represents the expansion of each channel. This process ensures that the model can flexibly focus on the key parts of the input features, thereby significantly improving its ability to focus on important features and overall performance. In addition, this module aims to optimize the feature selection process, reducing the computational complexity while maintaining high efficiency, so as to better segment weeds that occlude each other.

[0086] Since the growth direction of corn plants in the field may be tilted, this embodiment uses the improved YOLOv10 model to predict the rotating bounding box instead of the traditional horizontal border to frame the plant area, which can more accurately adapt to the growth direction of corn crops, not only considering the angle and orientation of the corn plant, but also more accurately delineating the boundaries of the plant. At the same time, the improved YOLOv10 model predicts key points to accurately mark the center position of the corn plant, effectively improving the accuracy of the uniformity assessment of the corn plant.

[0087] In order to enable the improved YOLOv10 model to accurately predict the rotation bounding box and key points corresponding to the corn plant image, the improved YOLOv10 model is first trained in this embodiment. The training method of the improved YOLOv10 model includes: obtaining a number of historical remote sensing images, annotating the rotation bounding box and key points of each of the historical remote sensing images, and using the annotated historical remote sensing images as training data, wherein the historical remote sensing images include various growth cycles of corn seedlings; constructing a loss function corresponding to the improved YOLOv10 model; training the improved YOLOv10 model according to the training data and the loss function to obtain a trained improved YOLOv10 model, thereby predicting the growth cycle, rotation bounding box and key points of the corn plant according to the improved YOLOv10 model.

[0088] In one implementation, constructing the loss function corresponding to the improved YOLOv10 model includes:

[0089] Calculating the classification loss, bounding box loss, and key point loss during the training of the improved YOLOv10 model;

[0090] The classification loss, the bounding box loss, and the key point loss are weighted to determine a loss function corresponding to the improved YOLOv10 model.

[0091] Since the improved YOLOv10 model is used to predict the rotated bounding box and key points corresponding to the corn plant image, and the corn plant and other elements need to be classified at the same time, the classification loss, bounding box loss and key point loss of the improved YOLOv10 model in this embodiment during the training process are weighted to obtain the loss function corresponding to the improved YOLOv10 model.

[0092] 1. Classification loss (ClassLoss) is binary cross entropy loss (BCE), the formula is as follows:

[0093]

[0094] y i is the class label of the corn plant, which is used to represent the growth cycle of the corn plant. i is the probability predicted by the improved YOLOv10 model, and N is the total number of corn plants.

[0095] 2. The bounding box loss (BboxLoss) includes the rotated bounding box loss (RotatedBboxLoss) and the distribution focal loss (DistributionFocalLoss, DFL).

[0096] The formula for the rotation bounding box loss is as follows:

[0097] L bbox =IoU rotated (B pred , B gt ),

[0098] Among them B pred is the rotation bounding box predicted by the improved YOLOv10 model, B gt is the true rotated bounding box, calculated using IoU or GIoU.

[0099] The calculation formula of distributed focal loss is as follows:

[0100] L dfl =Softmax(dist)·proj,

[0101] Where disf is the distribution predicted by the improved YOLOv10 model, and proj is a projection operation used to calculate the rotation bounding box parameters in the discrete distribution.

[0102] 3. Keypoint Loss includes keypoint position loss and keypoint object loss.

[0103] The calculation formula of key point position loss is as follows:

[0104]

[0105] in is the improved YOLOv10 model predicting the key point coordinates, u i is the real key point, A i is the target area, which is used to balance the importance of key points.

[0106] The formula for calculating the key point object loss is as follows:

[0107]

[0108] in is the keypoint visibility label predicted by the improved YOLOv10 model, h i is the true label, calculated using binary cross entropy.

[0109] 4. Weight the classification loss, bounding box loss, and key point loss to determine the loss function corresponding to the improved YOLOv10 model as follows:

[0110] L total =Lbbox ·w box +L kpts ·w pose +L kpts_obj ·w kobj +L cls ·w cls +L dfl ·w dfl ,

[0111] where w box YesL bbox The corresponding weight, w pose YesL kpts The corresponding weight, w kobj YesL kpts_obj The corresponding weight, w cls YesL cfs The corresponding weight, w dfl YesL dfl The corresponding weight.

[0112] This embodiment uses remote sensing images containing corn plants or disturbed plants of different sizes for testing. The results show that the improved YOLOv10 model can accurately predict the rotation bounding box and key points of corn plants in different situations, such as Figure 5-7 shown.

[0113] Step S300, determining the corn germination rate and germination uniformity corresponding to the target area according to the rotation boundary boxes and key points corresponding to each of the corn plant images.

[0114] In simple terms, each corn plant corresponds to a rotation bounding box, and the number of corn seedlings can be obtained according to the number of rotation bounding boxes, thereby calculating the corn seedling emergence rate. The key point represents the center or center of the corn plant, and the uniformity of corn seedling emergence can be obtained according to the uniformity of the key point.

[0115] In one implementation, determining the corn emergence rate corresponding to the target area according to the rotation bounding box corresponding to each of the corn plant images includes:

[0116] Step S301, determining the number of rotation frames corresponding to the target area according to the rotation boundary frames corresponding to the corn plant images;

[0117] Step S302: acquiring the area corresponding to the target area, and determining the corn emergence rate of the target area according to the number of rotating boxes and the area.

[0118] Specifically, this embodiment calculates the number of rotating boxes n corresponding to the target area according to the rotating boundary boxes corresponding to each corn plant image, and uses the number of rotating boxes to represent the number of corn seedlings. The actual area m corresponding to the target area or remote sensing image is obtained, and the corn emergence rate (Emergence rate) corresponding to the target area or remote sensing image is calculated according to the number of rotating boxes and the area:

[0119]

[0120] In one implementation, determining the uniformity of seedling emergence corresponding to the target area according to the key points corresponding to each of the corn plant images includes:

[0121] Step S303, determining the pixel distance between adjacent corn plants according to the key points corresponding to the corn plant images;

[0122] Step S304, obtaining a ground sampling distance, and determining the actual distance between adjacent corn plants according to each pixel distance and the ground sampling distance;

[0123] Step S305: determining the seedling emergence uniformity corresponding to the target area according to the actual distance between adjacent corn plants.

[0124] Specifically, this embodiment counts the pixel distances of adjacent corn plants according to the key points (centers of each corn) corresponding to each corn plant image, and calculates the actual ground area corresponding to the pixels in the remote sensing image through the ground sampling distance (GSD). The ground sampling distance represents the actual distance on the ground corresponding to a pixel in the remote sensing image. If the GSD is 1 cm, it means that each pixel in the image represents an area of ​​1 square centimeter on the ground.

[0125]

[0126] Where H is the flight altitude (distance from the ground), S is the pixel size of the sensor, and f is the focal length of the sensor. If the pixel distance between the centers (key points) of two corns is d, then the actual distance D between the two corns is:

[0127] D = d × GSD,

[0128] Determine the uniformity of seedling emergence in the target area based on the actual distance between adjacent corn plants, including:

[0129] Calculate the average value μ of the actual distance between adjacent corn plants. Assume that there are n pairs of actual distances between adjacent plants, which are recorded as D1, D2, ..., D n The average value represents the average level of the actual distance between adjacent plants, and the calculation formula is:

[0130]

[0131] Where μ is the mean value. i is the actual distance between the ith pair of adjacent corn plants. n is the number of adjacent corn plant pairs.

[0132] Calculate the standard deviation σ, which measures the degree of dispersion of these intervals. The calculation formula is:

[0133]

[0134] Where σ is the standard deviation of the actual distance between adjacent corn plants. i is the actual distance between the ith pair of adjacent corn plants. μ is the average value. n is the number of adjacent corn plant pairs.

[0135] Calculate the uniformity of seedling emergence in the target area:

[0136]

[0137] Where CV is the coefficient of variation, σ is the standard deviation, and μ is the mean.

[0138] In one implementation, this embodiment can be used to evaluate the emergence rate and uniformity of other crops, such as wheat, soybeans, etc., in addition to corn plants.

[0139] Based on the above embodiments, the present invention also provides a device for evaluating corn emergence rate and emergence uniformity, such as Figure 8 As shown, the device comprises:

[0140] Acquisition module 01, used to acquire a remote sensing image of a target area, and determine a number of corn plant images according to the remote sensing image;

[0141] A prediction module 02 is used to use an improved YOLOv10 model to make predictions based on the plurality of corn plant images, and determine a rotation bounding box and key points corresponding to each of the corn plant images, wherein the improved YOLOv10 model is implemented by introducing a Mamba attention module in a neck network layer of the YOLOv10 model, and the key points are used to mark the center of the corn plant;

[0142] The evaluation module 03 is used to determine the corn emergence rate and emergence uniformity corresponding to the target area according to the rotation boundary box and key points corresponding to each of the corn plant images.

[0143] Based on the above embodiment, the present invention further provides a terminal, whose principle block diagram can be shown as follows: Fig. 9As shown. The terminal includes a processor, a memory, a network interface, and a display screen connected through a system bus. Among them, the processor of the terminal is used to provide computing and control capabilities. The memory of the terminal includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the terminal is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for evaluating corn emergence rate and emergence uniformity is implemented. The display screen of the terminal can be a liquid crystal display screen or an electronic ink display screen.

[0144] Those skilled in the art will understand that Fig. 9 The principle block diagram shown in the figure is only a block diagram of a partial structure related to the scheme of the present invention, and does not constitute a limitation on the terminal to which the scheme of the present invention is applied. The specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0145] In one implementation, the terminal has one or more programs stored in its memory and is configured to be executed by one or more processors. The one or more programs include instructions for performing a method for evaluating corn emergence rate and emergence uniformity.

[0146] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0147] In summary, the present invention discloses a method, device, terminal and medium for evaluating corn emergence rate and emergence uniformity. The method determines a plurality of corn plant images according to a remote sensing image of a target area; uses an improved YOLOv10 model to predict a rotating bounding box and key points corresponding to each corn plant image, wherein the key point is the center of the corn plant; and determines the corn emergence rate and the emergence uniformity corresponding to the target area according to the rotating bounding box and key points corresponding to each corn plant image. Since the present invention predicts the rotating bounding box and key points of the corn plant by using an improved YOLOv10 model, the rotating bounding box can effectively reduce interference from background or other non-crop areas, improve detection accuracy, and the key points can accurately mark the center of the corn plant, and can more accurately evaluate the distance and distribution between plants. Therefore, the problem that the center of the corn plant is marked by the center point of the detected target box in the prior art, resulting in a large detection error, can be effectively solved.

[0148] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A method for evaluating corn germination rate and germination uniformity, characterized in that: The method comprises: Acquire a remote sensing image of a target area, and determine a plurality of corn plant images based on the remote sensing image; Using an improved YOLOv10 model to make predictions based on the plurality of corn plant images, determining a rotation bounding box and key points corresponding to each of the corn plant images, wherein the improved YOLOv10 model is implemented by introducing a Mamba attention module in a neck network layer of the YOLOv10 model, and the key points are used to mark the center of the corn plant; The corn germination rate and germination uniformity corresponding to the target area are determined according to the rotation bounding box and key points corresponding to each of the corn plant images.

2. The method for evaluating corn germination rate and germination uniformity according to claim 1, characterized in that: The step of determining a plurality of corn plant images according to the remote sensing image comprises: Extracting straight lines between rows of corn plants in the remote sensing image, and determining rows of corn plants in the remote sensing image based on the straight lines between rows; The remote sensing image is segmented according to the corn plant rows to determine a plurality of corn plant images.

3. The method for evaluating corn germination rate and germination uniformity according to claim 2, characterized in that: The step of extracting straight lines between rows of corn plants in the remote sensing image comprises: Performing image enhancement on the corn plants in the remote sensing image, and converting the enhanced remote sensing image into a binary image; Morphological operations and Hough transforms are performed on the binary image to extract the straight lines between the rows of corn plants in the remote sensing image.

4. The method for evaluating corn germination rate and germination uniformity according to claim 1, characterized in that: The training method of the improved YOLOv10 model includes: Acquire a number of historical remote sensing images, annotate each of the historical remote sensing images with a rotating bounding box and key points, and use the annotated historical remote sensing images as training data; Constructing a loss function corresponding to the improved YOLOv10 model; The improved YOLOv10 model is trained according to the training data and the loss function.

5. The method for evaluating corn germination rate and germination uniformity according to claim 4, characterized in that: The loss function corresponding to the improved YOLOv10 model is constructed, including: Calculating the classification loss, bounding box loss, and key point loss during the training of the improved YOLOv10 model; The classification loss, the bounding box loss, and the key point loss are weighted to determine a loss function corresponding to the improved YOLOv10 model.

6. The method for evaluating corn emergence rate and emergence uniformity according to claim 1, characterized in that: Determining the corn emergence rate corresponding to the target area according to the rotation bounding box corresponding to each of the corn plant images includes: Determining the number of rotation frames corresponding to the target area according to the rotation boundary frames corresponding to each of the corn plant images; The area of ​​the target area is obtained, and the corn emergence rate of the target area is determined according to the number of rotating boxes and the area of ​​the target area.

7. The method for evaluating corn emergence rate and emergence uniformity according to claim 1, characterized in that: Determining the uniformity of emergence of the seedlings corresponding to the target area according to the key points corresponding to each of the corn plant images includes: Determining the pixel distance between adjacent corn plants according to the key points corresponding to the corn plant images; Acquire a ground sampling distance, and determine the actual distance between adjacent corn plants according to each pixel distance and the ground sampling distance; The seedling emergence uniformity corresponding to the target area is determined according to the actual distance between each adjacent corn plant.

8. A device for evaluating corn emergence rate and uniformity, characterized in that: The device comprises: An acquisition module, used for acquiring a remote sensing image of a target area, and determining a plurality of corn plant images according to the remote sensing image; A prediction module, configured to use an improved YOLOv10 model to perform prediction based on the plurality of corn plant images, and determine a rotation bounding box and key points corresponding to each of the corn plant images, wherein the improved YOLOv10 model is implemented by introducing a Mamba attention module in a neck network layer of the YOLOv10 model, and the key points are used to mark the center of the corn plant; An evaluation module is used to determine the corn seedling emergence rate and seedling emergence uniformity corresponding to the target area according to the rotation boundary box and key points corresponding to each of the corn plant images.

9. A terminal, characterized in that: The terminal includes a memory and one or more processors; the memory stores one or more programs; the program contains instructions for executing the corn emergence rate and emergence uniformity evaluation method as described in any one of claims 1-7; and the processor is used to execute the program.

10. A computer-readable storage medium having a plurality of instructions stored thereon, characterized in that: The instructions are suitable for being loaded and executed by a processor to implement the steps of the method for evaluating corn emergence rate and emergence uniformity as described in any one of claims 1 to 7 above.

Images