A method for rapid ecological restoration of slope

Abstract

This invention provides a method for rapid ecological restoration of slopes. It utilizes neural networks for weed removal, improving the efficiency and accuracy of slope clearing. Furthermore, by mechanically breaking up the surface of the drilled area, the plant growth substrate can penetrate more comprehensively into the slope, thereby increasing the penetration range of the plant growth substrate and improving the efficiency of slope ecological restoration.

Description

Technical Field

[0001] This invention relates to the field of slope ecological restoration engineering technology, and in particular to a method for rapid ecological restoration of slopes. Background Technology

[0002] The canal slopes not only protect the banks from the scouring caused by the strong currents from ship waves and propellers, as well as the damage that can be caused by changes in water flow and water level in the canal, and the erosion caused by rainwater and groundwater, but also allow the vegetation planted on the canal slopes to fix carbon and release oxygen.

[0003] Currently, due to external factors, the original vegetation on the canal slopes has been severely damaged, resulting in a large number of bare soil or rock slopes. It is difficult for such canal slopes to restore their ecological balance by nature's own power, which has exacerbated the deterioration of the ecological environment. Therefore, the development and improvement of vegetation protection technology for canal slopes is of great significance to ecological environment construction.

[0004] Currently, most slope clearing methods rely on manual weeding. This approach cannot distinguish between existing seedlings and weeds, leading to damage to existing seedlings. After clearing, spraying is often done with mixed materials to achieve rapid ecological restoration of the canal slope, but the restoration speed is slow and cannot achieve an efficient slope restoration process. Summary of the Invention

[0005] In order to overcome the shortcomings of the existing technology, the purpose of this invention is to provide a method for rapid ecological restoration of slopes.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A method for rapid ecological restoration of slopes includes:

[0008] The target slope was surveyed on-site to obtain survey data and real-time on-site images; the on-site images included images of seedlings and weeds.

[0009] Based on the survey data, a structural model of the target slope is obtained;

[0010] Based on the trained neural network model, the seedling image and the weed image are identified and located to obtain the center pixel coordinates and width and height of the minimum detection box of the seedling and the center pixel coordinates of the minimum detection box of the weed.

[0011] The coordinates of the four corners of the seedling's minimum detection box are calculated based on the center pixel coordinates and the width and height of the minimum detection box.

[0012] Real-time slope crop row map information is constructed based on the center and corner coordinates of the minimum detection box of the seedlings and the center pixel coordinates of the minimum detection box of the weeds.

[0013] Determine weeding strategies based on the slope crop row map information;

[0014] The target slope is cleared according to the weeding strategy and the slope crop row map information;

[0015] Holes are drilled into the cleared target slope, and mechanical crushing is performed in the drilled area to break the surface of the drilled area.

[0016] The growth substrate was filled, and the filter material was laid and maintained in the broken area.

[0017] Preferably, the step of conducting on-site surveying of the target slope to obtain surveying data and real-time on-site images includes:

[0018] Based on on-site surveying and data collection, the structural data of the target slope were acquired, and the surveying data was obtained.

[0019] The surface of each area of ​​the target slope is photographed using a camera to obtain the on-site images.

[0020] Preferably, the method for constructing the neural network model includes:

[0021] Construct a convolutional neural network;

[0022] Based on images of crop seedlings and weeds on slopes from historical datasets, we determine the training and validation datasets for seedling images and the training and validation datasets for weed images.

[0023] The training dataset is input into the convolutional neural network for training to obtain a trained model, and the trained model is tested using the validation dataset to obtain a trained neural network model.

[0024] Preferably, real-time slope crop row map information is constructed based on the center coordinates and four corner coordinates of the minimum detection box of the seedlings and the center pixel coordinates of the minimum detection box of the weeds, including:

[0025] The work area is divided into a safe zone and a danger zone based on the center coordinates and four corner coordinates of the minimum detection frame of the seedling;

[0026] Based on the pixel coordinates of the safe and dangerous zones, a slope crop row map is constructed according to the center pixel coordinates of the smallest weed detection box.

[0027] Preferably, determining a weeding strategy based on the slope crop row map information includes:

[0028] Based on the slope crop row map information, determine whether the number of weeds in the work area is equal to zero, and obtain the first judgment result;

[0029] If the first judgment result is yes, then no weeding measures will be taken;

[0030] If the first judgment result is negative, then it is determined whether the number of weeds is greater than the threshold to obtain the second judgment result. If the second judgment result is positive, then it is determined that the weed density is high, and manual weeding is used for indiscriminate weeding.

[0031] If the second judgment result is negative, then calculate the angle between the line connecting adjacent weeds and the vertical direction to obtain the first angle and the second angle; wherein, the first angle is the angle between the line connecting the center points of adjacent weeds and the vertical direction, and the second angle is the angle between the line connecting the center points of the first weed in the safe area and the last weed in the danger area and the vertical direction.

[0032] Calculate the angle between the maximum lateral movement speed of the laser weeding device and the forward speed of the weeding equipment to obtain the third angle;

[0033] Determine if the third included angle is greater than the first or second included angle. If so, use laser mode for weeding; otherwise, use mechanical mode for weeding.

[0034] Preferably, holes are made in the cleared target slope, and mechanical crushing is performed in the area of ​​the holes to break the surface of the area where the holes are made, including:

[0035] Drill holes at an angle of 70 to 120 degrees to the horizontal plane on the exposed rock slope of the canal. The hole diameter is 4 to 8 cm, the hole depth is 40 to 60 cm, and the hole spacing is 20 to 40 cm.

[0036] Using rock-breaking tools, mechanical fracturing with a preset force is applied to each drilled area to break the surface of the drilled area.

[0037] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0038] This invention provides a method for rapid ecological restoration of slopes, comprising: conducting on-site surveying of the target slope to obtain surveying data and real-time on-site images; the on-site images include seedling images and weed images; modeling based on the surveying data to obtain a structural model of the target slope; based on a trained neural network model, identifying and locating the seedling images and the weed images to obtain the center pixel coordinates and width / height of the seedling's minimum detection frame and the center pixel coordinates of the weed's minimum detection frame; calculating the four corner coordinates of the seedling's minimum detection frame based on the center pixel coordinates and width / height of the seedling's minimum detection frame; constructing real-time slope crop row map information based on the center coordinates and four corner coordinates of the seedling's minimum detection frame and the center pixel coordinates of the weed's minimum detection frame; determining a weeding strategy based on the slope crop row map information; cleaning the target slope based on the weeding strategy and the slope crop row map information; making openings in the cleaned target slope and mechanically breaking the surface of the opening areas; filling the broken areas with growth substrate, laying drainage materials, and maintaining them. This invention utilizes neural networks for weed removal, improving the efficiency and precision of slope clearing. By mechanically breaking up the surface of the drilled area, the plant growth substrate can penetrate more comprehensively into the slope, thereby increasing the penetration range of the plant growth substrate and improving the efficiency of slope ecological restoration. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart of a method provided in an embodiment of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0043] The terms "first," "second," "third," and "fourth," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, including a series of steps, processes, methods, etc., is not limited to the steps listed, but may optionally include steps not listed, or may optionally include other steps inherent to these processes, methods, products, or devices.

[0044] The purpose of this invention is to provide a method for rapid ecological restoration of slopes. By mechanically breaking up the surface of the drilled area, the plant growth substrate can be more fully penetrated into the slope, thereby increasing the penetration range of the plant growth substrate and improving the efficiency of slope ecological restoration.

[0045] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0046] Figure 1 The method flowchart provided in the embodiments of the present invention is as follows: Figure 1 As shown, this invention provides a method for rapid ecological restoration of slopes, comprising:

[0047] A method for rapid ecological restoration of slopes includes:

[0048] Step 100: Conduct on-site surveying of the target slope to obtain surveying data and real-time on-site images; the on-site images include images of seedlings and weeds;

[0049] Step 200: Model the target slope based on the survey data to obtain its structural model;

[0050] Step 300: Based on the trained neural network model, identify and locate the seedling image and the weed image to obtain the center pixel coordinates and width and height of the minimum detection box of the seedling and the center pixel coordinates of the minimum detection box of the weed.

[0051] Step 400: Calculate the coordinates of the four corners of the minimum detection box of the seedling based on the center pixel coordinates and width and height of the minimum detection box of the seedling;

[0052] Step 500: Construct real-time slope crop row map information based on the center coordinates and four corner coordinates of the minimum detection box of the seedling and the center pixel coordinates of the minimum detection box of the weeds;

[0053] Step 600: Determine the weeding strategy based on the slope crop row map information;

[0054] Step 700: Clean the target slope according to the weeding strategy and the slope crop row map information;

[0055] Step 800: Drill holes in the cleaned target slope and mechanically crush the area where the holes are drilled to break the surface of the drilled area;

[0056] Step 900: Fill the broken area with growth substrate, lay and maintain the filter material.

[0057] Preferably, the step of conducting on-site surveying of the target slope to obtain surveying data and real-time on-site images includes:

[0058] Based on on-site surveying and data collection, the structural data of the target slope were acquired, and the surveying data was obtained.

[0059] The surface of each area of ​​the target slope is photographed using a camera to obtain the on-site images.

[0060] Specifically, in this embodiment, the on-site surveying can be conducted by investigating the geological report, geographical data, and architectural drawings of the target slope area to obtain relevant surveying data. On-site images are obtained through on-site photography by staff or real-time aerial photography using drones.

[0061] Preferably, the method for constructing the neural network model includes:

[0062] Construct a convolutional neural network;

[0063] Based on images of crop seedlings and weeds on slopes from historical datasets, we determine the training and validation datasets for seedling images and the training and validation datasets for weed images.

[0064] The training dataset is input into the convolutional neural network for training to obtain a trained model, and the trained model is tested using the validation dataset to obtain a trained neural network model.

[0065] Specifically, the convolutional neural network in this embodiment can be one of VGGNet, Inception, ResNet, DenseNet, FPN, and DetNet. Among them, the preferred neural network in this application is DetNet.

[0066] Furthermore, DetNet, also known as a deterministic network, refers to a private network built using network resources that is predictable, plannable, designable, verifiable, and possesses deterministic capabilities. These deterministic capabilities encompass deterministic latency, jitter, packet loss, bandwidth, path determination, connectivity, and reliability. In a deterministic network, the occurrence of critical events can be pre-planned and scheduled. Through specific planning of communication tasks between various subsystems, critical tasks are kept in a conflict-free state.

[0067] Preferably, real-time slope crop row map information is constructed based on the center coordinates and four corner coordinates of the minimum detection box of the seedlings and the center pixel coordinates of the minimum detection box of the weeds, including:

[0068] The work area is divided into a safe zone and a danger zone based on the center coordinates and four corner coordinates of the minimum detection frame of the seedling;

[0069] Based on the pixel coordinates of the safe and dangerous zones, a slope crop row map is constructed according to the center pixel coordinates of the smallest weed detection box.

[0070] Specifically, by determining the horizontal coordinates of pixels, this embodiment further divides the safe zone and the danger zone, including a left safe zone, a right safe zone, a left danger zone, and a right danger zone. That is, the entire detection frame is composed of the left safe zone, the right safe zone, the left danger zone, and the right danger zone.

[0071] Preferably, determining a weeding strategy based on the slope crop row map information includes:

[0072] Based on the slope crop row map information, determine whether the number of weeds in the work area is equal to zero, and obtain the first judgment result;

[0073] If the first judgment result is yes, then no weeding measures will be taken;

[0074] If the first judgment result is negative, then it is determined whether the number of weeds is greater than the threshold to obtain the second judgment result. If the second judgment result is positive, then it is determined that the weed density is high, and manual weeding is used for indiscriminate weeding.

[0075] If the second judgment result is negative, then calculate the angle between the line connecting adjacent weeds and the vertical direction to obtain the first angle and the second angle; wherein, the first angle is the angle between the line connecting the center points of adjacent weeds and the vertical direction, and the second angle is the angle between the line connecting the center points of the first weed in the safe area and the last weed in the danger area and the vertical direction.

[0076] Calculate the angle between the maximum lateral movement speed of the laser weeding device and the forward speed of the weeding equipment to obtain the third angle;

[0077] Determine if the third included angle is greater than the first or second included angle. If so, use laser mode for weeding; otherwise, use mechanical mode for weeding.

[0078] Alternatively, laser weeding methods can precisely locate the location of weeds, thereby accurately removing them from the area, while mechanical weeding methods can indiscriminately remove weeds from the area, enabling large-scale and rapid removal.

[0079] Preferably, holes are made in the cleared target slope, and mechanical crushing is performed in the area of ​​the holes to break the surface of the area where the holes are made, including:

[0080] Drill holes at an angle of 70 to 120 degrees to the horizontal plane on the exposed rock slope of the canal. The hole diameter is 4 to 8 cm, the hole depth is 40 to 60 cm, and the hole spacing is 20 to 40 cm.

[0081] Using rock-breaking tools, mechanical fracturing with a preset force is applied to each drilled area to break the surface of the drilled area.

[0082] In this embodiment, after clearing, holes are drilled at an angle of 70-120° to the horizontal plane on the exposed rock slope of the canal. The hole diameter is 4-8 cm, the hole depth is 40-60 cm, and the hole spacing is 20-40 cm. A rock-breaking tool is used to mechanically break the drilled areas with a preset force to break the surface of the drilled areas. Plant growth substrate is then filled into the mechanically broken areas. A 2-3 cm thick filter material is laid on the exposed rock surface. An ecological fence is set up on the filter material laid in the third step. The ecological fence is made of wooden ecological boards with a thickness of 2-3 cm. Plant growth substrate is sprayed twice into the ecological fence created in the fourth step. The first spraying is seedless plant growth substrate with a thickness of 5-10 cm. The second spraying is seed-containing plant growth substrate with a thickness of 1-2 cm. Maintenance involves watering twice a day, morning and evening, for 50-80 days.

[0083] Preferably, the outer surface of the ecological fence is provided with an ecological net, and the ecological net is sprayed with a plant growth substrate containing seeds.

[0084] Furthermore, the surface of the plant growth substrate sprayed in the ecological fence is covered with a layer of wire mesh. Ecological partitions, made of wood, are installed along the slope direction of the exposed rock slope within the ecological fence.

[0085] Furthermore, in this embodiment, the ecological partitions are arranged at uniform intervals, and their materials, in addition to wood, can also be iron partitions and rubber partitions.

[0086] Preferably, the seeds include wisteria, creeping privet, Euonymus fortunei, trumpet vine, climbing rose, clematis, and star jasmine.

[0087] Specifically, this embodiment does not limit the type of seed. That is, for drilling areas in different climate zones, the type of seed will be adaptively selected according to geological conditions and climate type.

[0088] Preferably, the seedless plant growth substrate comprises: 30-40 parts planting soil, 5-10 parts local loess, 5-10 parts kaolin, 2-5 parts bentonite, 5-15 parts peat, 2-6 parts minerals, 0.5-1 part water-retaining agent, 5-15 parts straw fiber, 5-10 parts sawdust, 0.1-0.2 parts trace elements, 1-3 parts slow-release fertilizer, 1-2 parts wood fiber, and 5-10 parts loam-like substrate agent.

[0089] Specifically, the optimal time for fertilization is during the season when temperature and humidity are most suitable for vegetation growth. The amount of fertilizer depends on the type of grass, soil texture, season, and vegetation growth. After seedling emergence (or 15 days after hydroseeding), apply 0.3-0.5% urea solution 1-2 times (2.5 g / m²) along with watering. Late autumn fertilization is essential. Use slow-release compound fertilizer at a rate of 6-7 g / m² to promote root growth, ensuring safe overwintering and extending the green period. Fertilization can be done via foliar spraying (combined with watering) or broadcasting. When fertilizing, pay attention to the concentration; the urea concentration should be controlled below 0.5%. When broadcasting, ensure even distribution and water immediately after fertilization to prevent burning the seedlings.

[0090] Preferably, the seed-containing plant growth substrate is composed of a uniform mixture of fibrous material, humus, planting soil, plant binder, fertilizer, and seeds; each cubic meter of seed-containing growth substrate contains 0.15-0.35 cubic meters of fibrous material, 0.4-0.8 cubic meters of humus, and 0.1-0.3 cubic meters of planting soil, with the total volume of each component being 1 cubic meter; then 2-5 kg ​​of plant binder, 15-35 kg of fertilizer, and 100-200 g of seeds are added.

[0091] Optionally, this embodiment also includes a reseeding process. Reseeding is carried out 7-20 days after the hydroseeding ends. The seeds are soaked and germinated 48-72 hours before reseeding, and then sown manually or with a hydroseeder. Later monitoring is conducted to monitor the organic matter content of the soil.

[0092] Effect evaluation:

[0093] ① Twenty days after construction, the coverage of Trachelospermum jasminoides reached over 86%, and other plant seeds have begun to germinate.

[0094] ② 33 days after construction, the trachelospermum was completely covered, other plants were growing well, and the surface of the substrate was not washed away by rainwater.

[0095] ③The vegetation can safely overwinter, turns green well, and has a good landscape effect.

[0096] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0097] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for rapid ecological restoration of slopes, characterized in that, include: The target slope was surveyed on-site to obtain survey data and real-time on-site images; the on-site images included images of seedlings and weeds. Based on the survey data, a structural model of the target slope is obtained; Based on the trained neural network model, the seedling image and the weed image are identified and located to obtain the center pixel coordinates and width and height of the minimum detection box of the seedling and the center pixel coordinates of the minimum detection box of the weed. The coordinates of the four corners of the seedling's minimum detection box are calculated based on the center pixel coordinates and the width and height of the minimum detection box. Real-time slope crop row map information is constructed based on the center and corner coordinates of the minimum detection box of the seedlings and the center pixel coordinates of the minimum detection box of the weeds. Weeding strategies are determined based on the slope crop row map information; The target slope is cleared according to the weeding strategy and the slope crop row map information; Holes are drilled into the cleared target slope, and mechanical crushing is performed in the drilled area to break the surface of the drilled area. The growth substrate was filled, and the filter material was laid and maintained in the broken area.

2. The rapid ecological restoration method for slopes according to claim 1, characterized in that, The process of conducting on-site surveying of the target slope to obtain surveying data and real-time on-site images includes: Based on on-site surveying and data collection, the structural data of the target slope were acquired, and the surveying data was obtained. The surface of each area of ​​the target slope is photographed using a camera to obtain the on-site images.

3. The rapid ecological restoration method for slopes according to claim 1, characterized in that, The method for constructing the neural network model includes: Constructing a convolutional neural network; Based on images of crop seedlings and weeds on slopes from historical datasets, we determine the training and validation datasets for seedling images and the training and validation datasets for weed images. The training dataset is input into the convolutional neural network for training to obtain a trained model, and the trained model is tested using the validation dataset to obtain a trained neural network model.

4. The rapid ecological restoration method for slopes according to claim 1, characterized in that, Based on the center and corner coordinates of the seedling's minimum detection box and the center pixel coordinates of the weed's minimum detection box, a real-time slope crop row map is constructed, including: The work area is divided into a safe zone and a danger zone based on the center coordinates and four corner coordinates of the minimum detection frame of the seedling; Based on the pixel coordinates of the safe and dangerous zones, a slope crop row map is constructed according to the center pixel coordinates of the smallest weed detection box.

5. The rapid ecological restoration method for slopes according to claim 4, characterized in that, Weed control strategies are determined based on the slope crop row map information, including: Based on the slope crop row map information, determine whether the number of weeds in the work area is equal to zero, and obtain the first judgment result; If the first judgment result is yes, then no weeding measures will be taken; If the first judgment result is negative, then it is determined whether the number of weeds is greater than the threshold to obtain the second judgment result. If the second judgment result is positive, then it is determined that the weed density is high, and manual weeding is used for indiscriminate weeding. If the second judgment result is negative, then calculate the angle between the line connecting adjacent weeds and the vertical direction to obtain the first angle and the second angle; wherein, the first angle is the angle between the line connecting the center points of adjacent weeds and the vertical direction, and the second angle is the angle between the line connecting the center points of the first weed in the safe area and the last weed in the danger area and the vertical direction. Calculate the angle between the maximum lateral movement speed of the laser weeding device and the forward speed of the weeding equipment to obtain the third angle; Determine if the third included angle is greater than the first or second included angle. If so, use laser mode for weeding; otherwise, use mechanical mode for weeding.

6. The rapid ecological restoration method for slopes according to claim 1, characterized in that, Drill holes in the cleared target slope and mechanically break the surface of the drilled area, including: Drill holes at an angle of 70-120° to the horizontal plane on the exposed rock slope of the canal. The hole diameter is 4-8 cm, the hole depth is 40-60 cm, and the hole spacing is 20-40 cm. Using rock-breaking tools, mechanical fracturing with a preset force is applied to each drilled area to break the surface of the drilled area.

Images