Forest resource surveying method based on unmanned aerial vehicle technology

Abstract

The invention relates to a forest resource exploration method based on an unmanned aerial vehicle technology in the field of unmanned aerial vehicle exploration, and the method comprises the steps: dividing a whole to-be-explored forest region into a plurality of primary regions, setting the coordinates of feature shooting points capable of shooting a whole secondary region, obtaining a planned flight path based on the feature shooting points, and carrying out the exploration. The overall high-angle shot image of each primary area is obtained by adopting a high-angle shot mode of a feature shooting point, so that the inspection efficiency is greatly improved, and a fixed-path piece-by-piece inspection mode adopted by traditional unmanned aerial vehicle inspection is replaced; meanwhile, when a clear image reflecting the characteristics of the primary area cannot be shot in the primary area, the unmanned aerial vehicle reduces the shooting height, further divides the primary area and re-plans a flight path, so that the influence of visibility and natural illumination intensity on the image shooting quality is reduced; the problem of low exploration quality caused by poor shooting quality of traditional unmanned aerial vehicle patrol is solved.

Description

Technical Field

[0001] The invention relates to a forest resource exploration method, in particular to a forest resource exploration method based on unmanned aerial vehicle technology and applied in the field of unmanned aerial vehicle exploration. Background Art

[0002] In forest resource exploration, the existing drone exploration method uses drones to shoot piece by piece. Although the piece-by-piece shooting method can cover the entire forest, it has obvious disadvantages. This method not only leads to a huge amount of captured data, which increases the difficulty of subsequent data processing and analysis, but also the drone needs to fly along a fixed long path, with long flight time, high energy consumption, and low overall exploration efficiency.

[0003] The existing patent with publication number CN111444774B discloses a forest resource monitoring method based on UAV aerial survey technology. The specific monitoring method steps are as follows: Step 1: obtain remote sensing images of large-scale forestry resources and perform preprocessing to obtain forestry resource information; Step 2: classify the forestry resource information obtained in step 1 and perform statistical analysis in small classes; Step 3: use remote sensing technology to measure the forest volume of a large-scale forestry resource in small classes as units and extract remote sensing data, and establish a regression model in combination with a small amount of sample plot survey data; Step 4: technicians operate drones to shoot images of each small class work point to determine high-definition image data. On the basis of previous research, combined with sample plot field survey data, the high-definition image is segmented at multiple scales. The invention can use drone image processing to obtain forestry resource information on a large scale. The detection method is simple and accurate, requires a short time, is highly practical, and is suitable for wide promotion and use.

[0004] The above patent improves exploration efficiency by combining remote sensing technology and drone technology to carry out zoning, photography and statistics, and takes images for cutting and further analysis. However, it does not solve the problem that the long flight path of drone exploration and the low quality of the images taken affect the exploration quality. Summary of the invention

[0005] In view of the above-mentioned prior art, the technical problem to be solved by the present invention is that the existing UAV has a long flight path when exploring forest resources and has a low image shooting quality, which affects the exploration quality.

[0006] To solve the above problems, the present invention provides a forest resource survey method based on UAV technology, comprising the following steps:

[0007] Step 1: Primary division of the entire forest area to be explored to obtain multiple primary areas and corresponding flight paths and characteristic shooting points;

[0008] Step 2: Based on the flight paths and characteristic shooting points of the primary division level in step 1, the drone flies to the characteristic shooting point closest to itself to take aerial photography with an optical camera to obtain an optical aerial image;

[0009] Step 3: Analyze the optical overhead image in step 2 to determine whether it is a clear image. If the optical overhead image is a clear image, execute step 4; if the optical overhead image is not a clear image, execute step 7.

[0010] Step 4: The drone flies to a feature shooting point in the next primary area based on the flight path of the primary division level in step 1 to take pictures, and obtain an optical overhead image;

[0011] Step 5, analyzing the optical overhead image in step 4 to determine whether it is a clear image. If the optical overhead image is a clear image, executing step 6; if the optical overhead image is not a clear image, executing step 7;

[0012] Step 6: Determine whether the shooting of all the primary areas divided in step 1 is completed. If the shooting of all the primary areas divided in step 1 is completed, the exploration is completed and the drone returns. If the shooting of all the primary areas divided in step 1 is not completed, execute step 4.

[0013] Step 7, lowering the vertical shooting height of the drone, and determining whether the lowered shooting height of the drone is higher than the set minimum optical shooting height. When the lowered shooting height of the drone is higher than the set minimum optical shooting height, executing step 8; when the lowered shooting height of the drone is not higher than the set minimum optical shooting height, executing step 17;

[0014] Step 8, the drone takes pictures at a lowered height to obtain an optical overhead image, and analyzes the optical overhead image. When the optical overhead image is a clear image, step 9 is executed; when the optical overhead image is not a clear image, step 7 is executed;

[0015] Step nine, re-dividing the divided area where the drone is currently located, obtaining multiple secondary divided areas, as well as corresponding flight paths and characteristic shooting points, and recording the division levels;

[0016] Step 10: The drone flies to the characteristic shooting point closest to itself based on the flight path of the current divided level in step 9 to shoot, and obtains an optical overhead image;

[0017] Step 11, analyzing the optical overhead image obtained in step 10 to determine whether it is a clear image, if the optical overhead image is a clear image, executing step 12, if the optical overhead image is not a clear image, executing step 7;

[0018] Step 12: The drone flies to the next feature shooting point based on the flight path of the current divided level to obtain an optical overhead image;

[0019] Step 13, analyzing the optical overhead image obtained in step 12 to determine whether it is a clear image. If the optical overhead image is a clear image, executing step 14; if the optical overhead image is not a clear image, executing step 7;

[0020] Step 14, determining whether all sub-regions of the current division level have completed shooting, if all sub-regions of the current division level have completed shooting, executing step 15, if all sub-regions of the current division level have not completed shooting, executing step 12;

[0021] Step 15, determining whether all sub-regions of all the division levels in step 9 have been photographed. If all the sub-regions of all the division levels in step 9 have been photographed, executing step 4. If all the sub-regions of all the division levels in step 9 have not been photographed, executing step 16.

[0022] Step 16, the drone returns to the flight path of the previous level of the current level, and then executes step 12;

[0023] Step 17: start the thermal infrared camera on the drone, divide the current divided area (primary area or secondary area) based on the shooting capability of the thermal infrared camera, obtain multiple secondary areas and corresponding flight paths and feature shooting points, and record the division levels;

[0024] Step 18, the drone photographs each secondary area one by one according to the hierarchical flight path in step 17 to obtain infrared images, and then executes step 15;

[0025] Among them, the drone is fixedly connected with an optical camera, a thermal infrared camera and a controller.

[0026] In the above-mentioned forest resource survey method based on UAV technology, the exploration area is divided into multiple levels and aerial photography is performed at characteristic shooting points to improve the quality of exploration shooting, optimize the flight path, and improve the exploration efficiency.

[0027] As a further improvement of the present application, the primary division of the entire forest area to be explored in step 1 includes the following sub-steps:

[0028] S1, obtaining the characteristic shooting height of the drone at the primary division level: based on the historical average values ​​of visibility and natural light intensity in the forest area to be explored and the shooting capability of the optical camera carried by the drone, the characteristic shooting height of the drone at the primary division level is set;

[0029] S2, obtaining each primary area at the primary division level and the characteristic shooting point of each primary area: determining the actual forest area that can be captured by the image taken by the optical camera on the drone when taking a bird's-eye view based on the characteristic shooting height of the drone, and then dividing the entire forest area to be explored into multiple primary areas based on the actual forest area captured by a single shot image, so that the area of ​​each primary area is smaller than the actual forest area captured by the image taken by the optical camera at the characteristic shooting height, and then obtaining the actual geographic coordinates of each primary area based on the actual geographic information of the primary area, and determining the spatial coordinates of the characteristic shooting point of the drone in each secondary area based on the actual geographic coordinates of each primary area and the characteristic shooting height data of the drone;

[0030] S3, based on the spatial coordinates of the characteristic shooting points of the UAV in each primary area, plans the shortest flight path for the UAV to complete all shooting.

[0031] As a further improvement of the present application, the division of regions in step nine includes the following sub-steps:

[0032] A1, the drone obtains the coordinates of the test shooting point after the drone is lowered based on the coordinates of the characteristic shooting points of the divided level and the vertical height value of the drone;

[0033] A2, take photos at the test shooting point and analyze the captured images;

[0034] A3, repeat A1 and A2 until the captured image is a clear image and the shooting height is higher than the minimum optical shooting height, use the test shooting point as a transition shooting point, and further divide the secondary area based on the actual forest area that can be captured by the image captured at the transition shooting point to obtain a new secondary area of ​​a new division level, as well as new feature shooting points and flight paths based on the new secondary area.

[0035] As a further improvement of the present application, the controller is equipped with a shooting control system, which includes a control module, the control module is connected to an optical shooting module, an infrared shooting module, an image analysis module, a terminal communication module and a flight control module, the optical shooting module is connected to the optical camera, the infrared shooting module is connected to the thermal infrared camera, the terminal communication module is connected to a terminal communicator fixedly connected to the drone, the terminal communicator is connected to the monitoring end communicator of the monitoring center via a wireless network, and the image analysis module analyzes the optical image and infrared image and determines the quality of the captured image.

[0036] As a further improvement of the present application, the optical camera is connected to the drone through an adjusting gimbal, the control module is also connected to a shooting angle adjustment module, the shooting angle adjustment module is connected to the adjusting gimbal, the control module is also connected to an early warning module, the early warning module is used to generate early warning information, the early warning information includes coordinate information of the fire location and prompt information indicating the existence of a fire; the control module is also connected to a manual command receiving module, and the manual command receiving module is connected to the communication module.

[0037] As a further improvement of the present application, the image analysis module includes an object feature extraction unit and a feature comparison unit. The object feature extraction unit extracts object features in the optically captured image and the infrared captured image to further identify the type of the object. The object feature comparison unit compares the object features of the currently captured image with the object features in the historically captured images to compare the differences in the objects captured in the forest at the same location.

[0038] As a further improvement of the present application, the drone is also fixedly connected to a visibility measuring instrument and a light intensity sensor, and the control module is also connected to a visibility monitoring module and a light intensity monitoring module, the visibility monitoring module is connected to the visibility measuring instrument, and the light intensity monitoring module is connected to the light sensor.

[0039] As a further improvement of the present application, a buzzer is also fixedly connected to the drone, and the buzzer is connected to the early warning module.

[0040] As a further improvement of the present application, a loudspeaker is also installed on the drone, and the control module is also connected to a personnel driving away module, which is connected to the loudspeaker.

[0041] In summary, the present invention divides the entire forest area to be explored into multiple primary areas, then sets the characteristic shooting height based on the historical average values ​​of visibility and natural light intensity of the forest area, and establishes the coordinates of the characteristic shooting points based on the characteristic shooting height, and finally obtains the primary flight path, instead of the traditional UAV patrol method of patrolling piece by piece along a fixed path, and adopts a method of aerial shooting at characteristic shooting points in multiple primary areas to obtain the overall image of the primary area, which greatly improves the efficiency of the inspection; at the same time, the image captured in each primary area is analyzed, and when it is impossible to capture a clear image, the shooting height is further lowered and the primary area is further divided to obtain new secondary areas and new characteristic shooting points, and the UAV patrol shooting path is further optimized according to actual natural conditions, and the quality of exploration and inspection is further improved, overcoming the problem of low patrol quality caused by visibility in traditional UAV patrols; in addition, when it is impossible to obtain an optically clear image of the secondary area, a thermal infrared camera is started instead of an optical camera to perform infrared photography of forest resources, and abnormal situations such as tree defects, fires, and illegal logging are discovered in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a schematic diagram of the exploration process for this application;

[0043] Figure 2 This is a module diagram of the shooting control system in this application;

[0044] Figure 3 This is a schematic diagram of the division of primary areas and secondary areas in this application;

[0045] Figure 4 A schematic diagram of the actual forest area captured by the characteristic shooting points and images in this application;

[0046] Figure 5 Schematic diagram of the flight path of the primary area and the flight path of the secondary area in this application. DETAILED DESCRIPTION

[0047] Two implementation modes of the present application are described in detail below with reference to the accompanying drawings.

[0048] The first implementation method:

[0049] Figure 1-5 A forest resource survey method based on drone technology is shown, comprising the following steps:

[0050] Step 1: Perform a primary division on the entire forest area to be explored to obtain multiple primary areas and corresponding flight paths and feature shooting points, including the following sub-steps:

[0051] S1, obtaining the characteristic shooting height of the drone at the primary division level: based on the historical average values ​​of visibility and natural light intensity in the forest area to be explored and the shooting capability of the optical camera carried by the drone, the characteristic shooting height of the drone at the primary division level is set;

[0052] For details, please refer to Figure 1 , Figure 3 and Figure 4 , the characteristic shooting height refers to the shooting height at which the camera carried by the drone can take a clear image, and a clear image refers to a shot image that can identify the characteristics of various objects (trees, wild animals, soil, etc.) in the forest after analysis; and the quality of the shot image is significantly affected by the local natural lighting conditions and visibility of the forest. Therefore, when planning the primary division level, it is necessary to refer to the historical average values ​​of visibility and natural light intensity in the forest area. In addition, the shooting capability of the optical camera carried by the drone obviously has a strong influence on the quality of the shot image, such as the zoom capability, wide-angle shooting capability, resolution, etc. of the optical camera. Therefore, it is necessary to comprehensively consider the selection of the characteristic shooting height, and the characteristic shooting height can be corrected after multiple actual inspections;

[0053] To further illustrate the process of setting the characteristic shooting height, an example is given below: Figure 4 First, obtain the historical average values ​​of visibility and natural light intensity in the local forest area by querying the geographic information website; second, select the weather with visibility and natural light intensity close to the historical average values ​​for patrol shooting test, start the drone to fly to the first set height, take pictures, and then analyze the quality of the captured images; third, continuously test different set shooting heights, record the shooting height range in which clear images can be taken, and then select the maximum shooting height within the shooting height range as the set characteristic shooting height of the area, because the larger the height value of the characteristic shooting height, the larger the actual forest area captured by the clear image, and the fewer times of shooting are required, which improves the efficiency of exploration shooting;

[0054] S2, obtaining each primary area at the primary division level and the characteristic shooting point of each primary area: determining the actual forest area that can be captured by the image taken by the optical camera on the drone when taking a bird's-eye view based on the characteristic shooting height of the drone, and then dividing the entire forest area to be explored into multiple primary areas based on the actual forest area captured by a single shot image, so that the area of ​​each primary area is smaller than the actual forest area captured by the image taken by the optical camera at the characteristic shooting height, and then obtaining the actual geographic coordinates of each primary area based on the actual geographic information of the primary area, and determining the spatial coordinates of the characteristic shooting point of the drone in each secondary area based on the actual geographic coordinates of each primary area and the characteristic shooting height data of the drone;

[0055] For details, please refer to Figure 4 The characteristic shooting point is generally selected above the geographical center of each divided area, and the actual longitude and latitude coordinates of the center of each primary area and the height value of the characteristic shooting height are combined as the spatial coordinates of the characteristic shooting point;

[0056] Also, see Figure 5 It should be noted that the shapes and sizes of the divided areas do not need to be the same. The shape of the entire forest area, the image quality, and the planned flight path need to be considered comprehensively. More specifically, when the entire forest area is an irregular shape, the divided area at the edge of the forest area is obviously a planned area with an irregular geometric shape.

[0057] S3, planning the shortest flight path for the drone to complete all shooting based on the spatial coordinates of the characteristic shooting points of the drone in each primary area;

[0058] Step 2: Based on the flight paths and characteristic shooting points of the primary division level in step 1, the drone flies to the characteristic shooting point closest to itself to take aerial photography with an optical camera to obtain an optical aerial image;

[0059] Step 3: Analyze the optical overhead image in step 2 to determine whether it is a clear image. If the optical overhead image is a clear image, execute step 4; if the optical overhead image is not a clear image, execute step 7.

[0060] Step 4: The drone flies to a feature shooting point in the next primary area based on the flight path of the primary division level in step 1 to take pictures, and obtain an optical overhead image;

[0061] Step 5, analyzing the optical overhead image in step 4 to determine whether it is a clear image. If the optical overhead image is a clear image, executing step 6; if the optical overhead image is not a clear image, executing step 7;

[0062] Step 6: Determine whether the shooting of all the primary areas divided in step 1 is completed. If the shooting of all the primary areas divided in step 1 is completed, the exploration is completed and the drone returns. If the shooting of all the primary areas divided in step 1 is not completed, execute step 4.

[0063] Specifically, through steps one to six, the drone is used to photograph and explore the primary area. To further illustrate the division and photography of the primary area, a specific example is given as follows. Assume that the entire area to be explored is divided into nine primary areas based on the historical average values ​​of local visibility and lighting conditions, and obtain the coordinates of the characteristic shooting points and the corresponding flight path of each primary area. Assume that the drone carries an optical camera and flies to the characteristic shooting points of the nine primary areas to obtain clear images. Then, the drone has completed the inspection and photography of the entire forest to be explored, including the nine secondary areas, in the current flight path. When a clear image cannot be obtained in a certain secondary area, the primary area needs to be divided.

[0064] Step 7, lowering the vertical shooting height of the drone, and determining whether the lowered shooting height of the drone is higher than the set minimum optical shooting height. When the lowered shooting height of the drone is higher than the set minimum optical shooting height, executing step 8; when the lowered shooting height of the drone is not higher than the set minimum optical shooting height, executing step 17;

[0065] Step 8, the drone takes pictures at a lowered height to obtain an optical overhead image, and analyzes the optical overhead image. When the optical overhead image is a clear image, step 9 is executed; when the optical overhead image is not a clear image, step 7 is executed;

[0066] Step nine, re-dividing the divided area where the drone is currently located, obtaining multiple secondary divided areas, as well as corresponding flight paths and characteristic shooting points, and recording the division levels;

[0067] For details, please refer to Figure 5 In step nine, the division of regions includes the following sub-steps:

[0068] A1, the drone obtains the coordinates of the test shooting point after the drone is lowered based on the coordinates of the characteristic shooting points of the divided level and the vertical height value of the drone;

[0069] A2, take photos at the test shooting point and analyze the captured images;

[0070] A3, repeat A1 and A2 until the captured image is a clear image and the shooting height is higher than the minimum optical shooting height, use the test shooting point as a transition shooting point, and then further divide the secondary area by the actual forest area that can be captured by the image captured at the transition shooting point, and obtain a new secondary area of ​​a new division level, as well as a new characteristic shooting point and flight path based on the new secondary area;

[0071] Among them, the vertical height of the drone lowered once is considered as a set value; in addition, the divided area in step nine includes the secondary area of ​​the primary area and the secondary areas of each divided level;

[0072] Step 10: The drone flies to the characteristic shooting point closest to itself based on the flight path of the current divided level in step 9 to shoot, and obtains an optical overhead image;

[0073] Step 11, analyzing the optical overhead image obtained in step 10 to determine whether it is a clear image, if the optical overhead image is a clear image, executing step 12, if the optical overhead image is not a clear image, executing step 7;

[0074] Step 12: The drone flies to the next feature shooting point based on the flight path of the current divided level to obtain an optical overhead image;

[0075] Step 13, analyzing the optical overhead image obtained in step 12 to determine whether it is a clear image. If the optical overhead image is a clear image, executing step 14; if the optical overhead image is not a clear image, executing step 7;

[0076] Step 14, determining whether all sub-regions of the current division level have completed shooting, if all sub-regions of the current division level have completed shooting, executing step 15, if all sub-regions of the current division level have not completed shooting, executing step 12;

[0077] Step 15, determining whether all sub-regions of all the division levels in step 9 have been photographed. If all the sub-regions of all the division levels in step 9 have been photographed, executing step 4. If all the sub-regions of all the division levels in step 9 have not been photographed, executing step 16.

[0078] Step 16, the drone returns to the flight path of the previous level of the current level, and then executes step 12;

[0079] Specifically, in step 14 and step 15, when the drone completes shooting of all the secondary areas of the divided levels, then the drone flies back to the flight path of the primary divided level to shoot the next characteristic shooting point of the primary area; when the secondary areas of the current divided level have not completed shooting, the drone flies to the next characteristic shooting point of the flight path of the current divided level to shoot; when all the secondary areas of the current divided level have completed shooting and the secondary areas of all the divided levels have not completed shooting, the drone flies back to the next characteristic shooting point on the flight path of the previous divided level to shoot, until all the secondary areas of all the divided levels have completed shooting;

[0080] Step 17: start the thermal infrared camera on the drone, divide the current divided area (primary area or secondary area) based on the shooting capability of the thermal infrared camera, obtain multiple secondary areas and corresponding flight paths and feature shooting points, and record the division levels;

[0081] Specifically, when the shooting height of the optical camera is lower than the set minimum shooting height, a thermal infrared camera is needed for infrared shooting (the minimum shooting height needs to take into account the collision with trees caused by the low altitude of the drone and the excessive number of secondary areas divided due to too much shooting point data caused by too low altitude, which does not meet the actual exploration needs). Since the thermal infrared camera is less affected by visibility and natural lighting conditions, although it cannot accurately obtain the object features of different objects in the forest, it can capture the fire heat area and the forest area with abnormal changes (missing trees, water diversion), as a supplement when the optical camera cannot patrol, and it also has a good early warning function;

[0082] It should be noted that, in general (without considering the zoom capability of the optical camera and the infrared enhancer of the thermal infrared camera), the shooting range of the thermal infrared camera is smaller than that of the optical camera. Therefore, the best shooting point of the thermal infrared camera in the division area of ​​the current division level is generally selected in combination with the area of ​​the current division area and the shooting capability of the thermal infrared camera. The current division area is divided again based on the actual forest area that can be captured by the best shooting point of the thermal infrared camera, and multiple secondary areas and corresponding flight paths and feature shooting points are obtained. More specifically, the selection of the best shooting point includes the following situations: In case one, the actual forest area captured by the clear image taken by the thermal infrared camera at the best shooting point is much larger than the actual area of ​​the secondary area of ​​the division level, then it is necessary to lower the height of the thermal infrared camera so that the actual forest area captured by the image is adapted to the area of ​​the secondary area of ​​the planned level, thereby improving the quality of the infrared image; In case two, when the actual forest area captured by the image taken by the thermal infrared camera at the best shooting point is smaller than the area of ​​the secondary area of ​​the division level, it is necessary to divide the division area of ​​the division level again to obtain a new secondary area, and then obtain a new feature shooting point and a new planned flight path;

[0083] Step 18, the drone photographs each secondary area one by one according to the hierarchical flight path in step 17 to obtain infrared images, and then executes step 15;

[0084] Specifically, through steps 7 to 18, the primary area and the secondary area are divided into multiple levels. To further illustrate the process of multi-level division, an example is given below: when the drone cannot obtain a clear image at a feature shooting point in a primary area, the primary area is divided into the first level to obtain multiple first-level secondary areas and corresponding flight paths (the shortest flight path passing through the feature shooting points of each first-level secondary area) and feature shooting points; and when a clear image cannot be obtained at a feature shooting point in a first-level secondary area, the first-level secondary area is divided into the second level to obtain multiple first-level secondary areas. Second-level sub-areas within the sub-areas and the corresponding flight paths and characteristic shooting points; when a characteristic shooting point of a second-level sub-area cannot obtain a clear image, the second-level sub-area is divided into a third level to obtain multiple third-level sub-areas, flight paths and characteristic shooting points located within the second-level sub-area, and the above process is repeated; when a characteristic shooting point of an N-th level sub-area cannot obtain a clear image, the N-th level sub-area is divided into an N+1th level to obtain multiple N+1th sub-areas, flight paths and characteristic shooting points located within the N-th level area;

[0085] At the same time, in the process of dividing the Nth level division level, it is necessary to consider whether the shooting height of the drone after being lowered is higher than the set minimum optical shooting height. When the shooting height is not higher than the set minimum optical shooting height, the search for the optical feature shooting point of the level division level is abandoned, and thermal infrared shooting is adopted. Based on the shooting capability of the thermal infrared camera, the sub-region of this level is divided, and multiple N+1th level sub-regions based on infrared shooting are obtained. The thermal infrared camera is used to shoot the multiple N+1th level sub-regions divided based on infrared shooting one by one (infrared shooting is less affected by visibility and light intensity than optical shooting, and the multiple N+1th level sub-regions of infrared shooting are not further divided). Then, after completing the infrared shooting of multiple N+1th level sub-regions, return to the flight path of the Nth level division level, and optically shoot the shooting points of the next Nth level sub-region;

[0086] Among them, the drone is fixedly connected with an optical camera, a thermal infrared camera and a controller.

[0087] Compared with the traditional forest resource survey method, the present invention divides the entire forest area to be explored into multiple primary secondary areas, then sets the characteristic shooting height based on the historical average values ​​of the visibility and natural light intensity of the forest area, and establishes the coordinates of the characteristic shooting points based on the characteristic shooting height, and finally obtains the primary planned flight path, instead of the traditional UAV patrol method of patrolling piece by piece along a fixed path, adopts a method of aerial shooting at characteristic shooting points in multiple secondary areas to obtain the overall image of the secondary area, which greatly improves the efficiency of the inspection; at the same time, the image captured in each secondary area is analyzed, and when it is impossible to capture a clear image, the shooting height is further reduced and the secondary area is further divided to obtain new secondary areas and new characteristic shooting points, and the UAV patrol shooting path is further optimized according to the actual natural conditions, the quality of exploration and inspection is further improved, and the problem of low patrol quality caused by visibility in traditional UAV patrol is overcome; in addition, when it is impossible to obtain a clear infrared image of the secondary area, a thermal infrared camera is started instead of an optical camera to perform infrared photography of forest resources, and abnormal situations such as tree defects, fires, and illegal logging are discovered in time.

[0088] See also Figure 2 The controller is equipped with a shooting control system, which includes a control module. The control module is connected to an optical shooting module, an infrared shooting module, an image analysis module, a terminal communication module and a flight control module. The optical shooting module is connected to an optical camera, the infrared shooting module is connected to a thermal infrared camera, the terminal communication module is connected to a terminal communicator fixedly connected to the UAV, and the terminal communicator is connected to a monitoring terminal communicator of a monitoring center via a wireless network. The image analysis module analyzes the optical image and the infrared image and determines the quality of the shot image.

[0089] See also Figure 2 The image analysis module includes an object feature extraction unit and a feature comparison unit. The object feature extraction unit extracts the object features in the optically photographed image and the infrared photographed image, and then identifies the type of the object. The object feature comparison unit compares the object features of the currently photographed image with the object features in the historical photographed images, and compares the differences between the photographed objects in the forest at the same location.

[0090] Specifically, the optical image and the infrared image are analyzed by the image analysis module to judge the quality of the captured image, and then the characteristic shooting height is adjusted in a targeted manner. At the same time, it is easy to detect the changes of objects in the same area of ​​the forest, and then to judge whether there is damage to the trees or illegal logging.

[0091] See also Figure 2 The drone is also fixedly connected with a visibility measuring instrument and a light intensity sensor, and the control module is also connected with a visibility monitoring module and a light intensity monitoring module. The visibility monitoring module is connected with the visibility measuring instrument, and the light intensity monitoring module is connected with the light sensor.

[0092] Specifically, the visibility and lighting conditions of the patrol area can be monitored in real time through the visibility meter and light intensity sensor fixed on the drone, which facilitates better adjustment of the drone's feature shooting height, and then better division of secondary areas, thereby improving the quality of patrol shooting images.

[0093] Second implementation method:

[0094] Figure 2 A forest resource survey method based on UAV technology is shown. On the basis of the first implementation mode, an optical camera is connected to the UAV through an adjustable gimbal, a control module is further connected to a shooting angle adjustment module, and the shooting angle adjustment module is connected to the adjustable gimbal. When the UAV flies from one shooting point to another, the sky above the forest area to be inspected is photographed, and then the photographed image is analyzed. When the image analysis module extracts the image feature of burning smoke in the photographed image, the planned flight path of the UAV is changed so that the UAV flies to the sky above the location where the smoke is located, and then makes a circular rotation to shoot, and transmits the on-site image back to the monitoring center and sends an early warning message to the monitoring center; when no burning smoke feature is found in the analysis of the photographed image, the flight continues according to the planned flight path.

[0095] Specifically, during the inspection process, images of the sky above the forest area to be inspected are taken, and the characteristics of the burning smoke in the images are identified. The real-time images are sent back to the burning site in a timely manner, so that the monitoring center personnel can take countermeasures and improve the timeliness of the early warning.

[0096] It should be noted that the adjustable gimbal is an angle adjustment device fixedly mounted on the drone, which enables the optical camera to shoot at various angles. This is a prior art and will not be described in detail in this application.

[0097] See also Figure 2 The control module is also connected to an early warning module, which is used to generate early warning information, including coordinate information of the fire location and prompt information indicating the existence of a fire; the control module is also connected to a manual command receiving module, which is connected to the communication module.

[0098] Specifically, when the personnel in the monitoring center need to operate the drone, they send a message to the manual command receiving module, and the manual command receiving module transmits the manual command to the control module, thereby controlling the operation of the drone.

[0099] See also Figure 2 ,A buzzer is also fixedly connected to the drone, and the buzzer is connected to the ,early warning module.

[0100] Specifically, when a fire occurs, the drone flies at the edge of the burning area and then activates a buzzer to alert firefighters and drive away animals, thereby improving the firefighting effect.

[0101] See also Figure 2 The drone is also equipped with a loudspeaker, and the control module is also connected to a personnel driving away module, which is connected to the loudspeaker. When the image analysis module compares the current captured image with the historical image and finds that the trees have fallen, the drone changes the planned flight path, approaches the fallen position of the trees, and circles over them, and turns on the optical camera and thermal infrared camera to continuously shoot the fallen area; when abnormal unidentified personnel are found in the fallen area, the loudspeaker is turned on to drive the personnel away; when no personnel are found, the area is photographed and transmitted back to the monitoring center, and the drone continues to patrol along the planned path.

[0102] Specifically, the control module is also connected to a personnel identity storage module to store the facial features of the staff in the forest area, and then determine whether the person is an abnormal unidentified person, thereby reducing the probability of driving away the wrong person.

[0103] In view of current practical needs, the above-mentioned implementation mode adopted in this application is not limited to the scope of protection. Various changes made within the knowledge scope of technical personnel in this field without departing from the concept of this application still fall within the scope of protection of the present invention.

Claims

1. A forest resource survey method based on drone technology, characterized in that: The steps include: Step 1: Primary division of the entire forest area to be explored to obtain multiple primary areas and corresponding flight paths and characteristic shooting points; Step 2: Based on the flight paths and characteristic shooting points of the primary division level in step 1, the drone flies to the characteristic shooting point closest to itself to take aerial photography with an optical camera to obtain an optical aerial image; Step 3: Analyze the optical overhead image in step 2 to determine whether it is a clear image. If the optical overhead image is a clear image, execute step 4; if the optical overhead image is not a clear image, execute step 7. Step 4: The drone flies to a feature shooting point in the next primary area based on the flight path of the primary division level in step 1 to take pictures, and obtain an optical overhead image; Step 5, analyzing the optical overhead image in step 4 to determine whether it is a clear image. If the optical overhead image is a clear image, executing step 6; if the optical overhead image is not a clear image, executing step 7; Step 6: Determine whether the shooting of all the primary areas divided in step 1 is completed. If the shooting of all the primary areas divided in step 1 is completed, the exploration is completed and the drone returns. If the shooting of all the primary areas divided in step 1 is not completed, execute step 4. Step 7, lowering the vertical shooting height of the drone, and determining whether the lowered shooting height of the drone is higher than the set minimum optical shooting height. When the lowered shooting height of the drone is higher than the set minimum optical shooting height, executing step 8; when the lowered shooting height of the drone is not higher than the set minimum optical shooting height, executing step 17; Step 8, the drone takes pictures at a lowered height to obtain an optical overhead image, and analyzes the optical overhead image. When the optical overhead image is a clear image, step 9 is executed; when the optical overhead image is not a clear image, step 7 is executed; Step nine, re-dividing the divided area where the drone is currently located, obtaining multiple secondary divided areas, as well as corresponding flight paths and characteristic shooting points, and recording the division levels; Step 10: The drone flies to the characteristic shooting point closest to itself based on the flight path of the current divided level in step 9 to shoot, and obtains an optical overhead image; Step 11, analyzing the optical overhead image obtained in step 10 to determine whether it is a clear image, if the optical overhead image is a clear image, executing step 12, if the optical overhead image is not a clear image, executing step 7; Step 12: The drone flies to the next feature shooting point based on the flight path of the current divided level to obtain an optical overhead image; Step 13, analyzing the optical overhead image obtained in step 12 to determine whether it is a clear image. If the optical overhead image is a clear image, executing step 14; if the optical overhead image is not a clear image, executing step 7; Step 14, determining whether all sub-regions of the current division level have completed shooting, if all sub-regions of the current division level have completed shooting, executing step 15, if all sub-regions of the current division level have not completed shooting, executing step 12; Step 15, determining whether all sub-regions of all the division levels in step 9 have been photographed. If all the sub-regions of all the division levels in step 9 have been photographed, executing step 4. If all the sub-regions of all the division levels in step 9 have not been photographed, executing step 16. Step 16, the drone returns to the flight path of the previous level of the current level, and then executes step 12; Step 17: start the thermal infrared camera on the drone, divide the current divided area (primary area or secondary area) based on the shooting capability of the thermal infrared camera, obtain multiple secondary areas and corresponding flight paths and feature shooting points, and record the division levels; Step 18, the drone photographs each secondary area one by one according to the hierarchical flight path in step 17 to obtain infrared images, and then executes step 15; Among them, the drone is fixedly connected with an optical camera, a thermal infrared camera and a controller.

2. The forest resource survey method based on drone technology according to claim 1 is characterized in that: In step 1, the primary division of the entire forest area to be explored includes the following sub-steps: S1, obtaining the characteristic shooting height of the drone at the primary division level: based on the historical average values ​​of visibility and natural light intensity in the forest area to be explored and the shooting capability of the optical camera carried by the drone, the characteristic shooting height of the drone at the primary division level is set; S2, obtaining each primary area at the primary division level and the characteristic shooting point of each primary area: determining the actual forest area that can be captured by the image taken by the optical camera on the drone when taking a bird's-eye view based on the characteristic shooting height of the drone, and then dividing the entire forest area to be explored into multiple primary areas based on the actual forest area captured by a single shot image, so that the area of ​​each primary area is smaller than the actual forest area captured by the image taken by the optical camera at the characteristic shooting height, and then obtaining the actual geographic coordinates of each primary area based on the actual geographic information of the primary area, and determining the spatial coordinates of the characteristic shooting point of the drone in each secondary area based on the actual geographic coordinates of each primary area and the characteristic shooting height data of the drone; S3, based on the spatial coordinates of the characteristic shooting points of the UAV in each primary area, plans the shortest flight path for the UAV to complete all shooting.

3. The forest resource survey method based on drone technology according to claim 1 is characterized in that: The division of regions in step nine includes the following sub-steps: A1, the drone obtains the coordinates of the test shooting point after the drone is lowered based on the coordinates of the characteristic shooting points of the divided level and the vertical height value of the drone; A2, take photos at the test shooting point and analyze the captured images; A3, repeat A1 and A2 until the captured image is a clear image and the shooting height is higher than the minimum optical shooting height, use the test shooting point as a transition shooting point, and further divide the secondary area based on the actual forest area that can be captured by the image captured at the transition shooting point to obtain a new secondary area of ​​a new division level, as well as new feature shooting points and flight paths based on the new secondary area.

4. The method for forest resource survey based on drone technology according to claim 1 is characterized in that: The controller is equipped with a shooting control system, which includes a control module, which is connected to an optical shooting module, an infrared shooting module, an image analysis module, a terminal communication module and a flight control module. The optical shooting module is connected to an optical camera, the infrared shooting module is connected to a thermal infrared camera, the terminal communication module is connected to a terminal communicator fixedly connected to the drone, and the terminal communicator is connected to a monitoring terminal communicator of a monitoring center via a wireless network. The image analysis module analyzes the optical image and the infrared image and determines the quality of the shot image.

5. The method for forest resource survey based on drone technology according to claim 4 is characterized in that: The optical camera is connected to the drone via an adjustable gimbal, the control module is also connected to a shooting angle adjustment module, the shooting angle adjustment module is connected to the adjustable gimbal, the control module is also connected to an early warning module, the early warning module is used to generate early warning information, the early warning information includes coordinate information of the fire location and prompt information indicating the existence of a fire; the control module is also connected to a manual command receiving module, the manual command receiving module is connected to the communication module.

6. The method for forest resource survey based on drone technology according to claim 4 is characterized in that: The image analysis module includes an object feature extraction unit and a feature comparison unit. The object feature extraction unit extracts object features from the optically captured image and the infrared captured image to identify the type of the object. The object feature comparison unit compares the object features of the currently captured image with the object features in the historical captured images to compare the differences between the objects captured in the forest at the same location.

7. The method for forest resource survey based on drone technology according to claim 4 is characterized in that: The UAV is also fixedly connected with a visibility measuring instrument and a light intensity sensor, and the control module is also connected with a visibility monitoring module and a light intensity monitoring module, the visibility monitoring module is connected with the visibility measuring instrument, and the light intensity monitoring module is connected with the light sensor.

8. The method for forest resource survey based on UAV technology according to claim 5 is characterized in that: The drone is also fixedly connected with a buzzer, which is connected to the early warning module.

9. The method for forest resource survey based on drone technology according to claim 8, characterized in that: A loudspeaker is also installed on the drone, and the control module is also connected to a personnel-expelling module, which is connected to the loudspeaker.

Images