Goldfish position identification method based on unmanned aerial vehicle cruise monitoring image

Abstract

The invention discloses an unmanned aerial vehicle cruise monitoring image-based cowfish position identification method. Two position sensors are arranged in the vertical direction of the axis of an unmanned aerial vehicle; according to detection data of the two position sensors, fitting to obtain a route trajectory curve of the unmanned aerial vehicle in the horizontal direction and a height trajectory curve of the unmanned aerial vehicle in the vertical direction, and according to a shooting moment T of a cruise monitoring image of the unmanned aerial vehicle, obtaining a horizontal position coordinate (X, Y) and height information Z of the unmanned aerial vehicle at the moment; the yaw angle of the unmanned aerial vehicle is calculated according to the route trajectory curve, the pitch angle of the unmanned aerial vehicle is calculated according to the height trajectory curve, and the roll angle of the unmanned aerial vehicle is calculated according to the relative position information of the two position sensors; and calculating position coordinates of the dolphin in a cruise monitoring image of the unmanned aerial vehicle, calculating position information corresponding to the shooting moment T of the dolphin in a geographic coordinate system by combining a yaw angle, a pitch angle and a roll angle of the unmanned aerial vehicle, and converting the position information into latitude and longitude information to finish the position identification of the dolphin.

Description

Technical Field

[0001] The present invention relates to the technical field of biological monitoring, and in particular to a method for identifying the position of a finless porpoise based on unmanned aerial vehicle cruise monitoring images. Background Art

[0002] With the increasing awareness of environmental protection and the development of science and technology, the protection and management of biological habitats have received increasing attention. Among them, the identification and monitoring of the habitats of whales such as the finless porpoise is an important part of protecting biodiversity. Accurate habitat prediction can help formulate protection measures and promote the survival and reproduction of the finless porpoise.

[0003] With the continuous maturity and development of drone technology, the use of drone cruise images for the identification and monitoring of whale habitats can achieve large-scale, high-resolution image acquisition, while reducing the impact on the ecological behavior of whales. Because whales move in the water in a variety of random ways, drones take aerial photos of finless porpoises, and the calculation of the position of finless porpoises in the image is very complicated. In the past, it was simply replaced by the drone position or the center of the photo, but the deviation was large, up to 50m, which was not conducive to judging the distance relationship between multiple finless porpoises. Summary of the invention

[0004] The present invention provides a method for identifying the position of a finless porpoise based on an unmanned aerial vehicle cruise monitoring image, which solves the technical problem in the prior art that the position of the finless porpoise is simply replaced by the position of the unmanned aerial vehicle, resulting in a large deviation.

[0005] The present invention can be achieved through the following technical solutions:

[0006] A method for identifying the position of a finless porpoise based on an unmanned aerial vehicle cruise monitoring image comprises the following steps:

[0007] Step 1: Two position sensors are arranged in the vertical direction of the axis of the drone;

[0008] Step 2: According to the detection data of the two position sensors, the horizontal route trajectory curve and the vertical height trajectory curve of the UAV are fitted, and then according to the shooting time T of the UAV cruise monitoring image, the horizontal position coordinates (X, Y) and height information Z of the UAV at this time are obtained;

[0009] Step 3: Calculate the yaw angle κ of the drone based on the route trajectory curve, calculate the pitch angle ω of the drone based on the altitude trajectory curve, and calculate the roll angle of the drone based on the relative position information of the two position sensors.

[0010] Step 4: Calculate the position coordinates (m, n) of the finless porpoise in the drone cruise monitoring image, and then combine the yaw angle k, pitch angle ω and roll angle of the drone The location information of the finless porpoise corresponding to the shooting time T in the geographic coordinate system is calculated and converted into longitude and latitude information to complete the location identification of the finless porpoise.

[0011] Furthermore, in step 3, the angle between the tangent direction of the route trajectory curve at the shooting time T and the axis direction of the aircraft is taken as the yaw angle κ of the UAV;

[0012] The angle between the tangent direction of the height trajectory curve at the shooting time T and the horizontal direction is taken as the pitch angle ω of the UAV;

[0013] Calculate the roll angle of the drone based on the distance D and relative height difference h between the two position sensors

[0014] Furthermore, in the step 2, detection data corresponding to two position sensors at multiple consecutive detection moments are selected, and the cruise monitoring image corresponding to the detection data contains finless porpoises. The average value of the horizontal position and the average value of the height information in the detection data are used as fitting data, and the least squares method is used to obtain the route trajectory curve and the height trajectory curve.

[0015] Furthermore, the image coordinate system is established with the center point of the cruise monitoring image as the coordinate origin, the forward direction of the UAV passing through the coordinate origin as the y-axis, and the right direction pointing to the flight direction through the coordinate origin as the x-axis, and the position coordinates (m, n) of the finless porpoise in the image coordinate system are calculated;

[0016] The following equation is used to calculate the position information (X j ,Y j ).

[0017]

[0018]

[0019] b1=cosωsinκ

[0020] b2=cosωcosκ

[0021] b3=-sinω

[0022]

[0023] Further, the position information of the finless porpoise (X j ,Y j ) to latitude and longitude (B,L)

[0024]

[0025] Among them, B needs to be calculated iteratively, N is the radius of curvature of the earth's angular circle, and e is the first eccentricity of the earth.

[0026] Furthermore, in step one, two position sensors are arranged on the UAV using a cross-shaped bracket, wherein the vertical rod of the cross-shaped bracket is arranged along the axis of the UAV, and the two position sensors are arranged at the ends of the horizontal rod of the cross-shaped bracket.

[0027] Furthermore, the position sensor adopts an RTK locator, and the drone is provided with a camera for capturing images of the finless porpoise cruising and monitoring in the water.

[0028] The beneficial technical effects of the present invention are:

[0029] By adding two GNSS positioning devices, the instantaneous posture of the drone is calculated, and the principle of photogrammetry is used to convert the coordinates to obtain the accurate position of the finless porpoise, thereby correcting the deviation of using the drone coordinates as the finless porpoise position coordinates or the photo center coordinates as the finless porpoise position coordinates. Therefore, the method proposed in the present invention can accurately calculate the position of the finless porpoise based on the position, heading information, posture information, etc. of the drone, which is conducive to statistical analysis of the position of the finless porpoise, and to determine the distance information between the finless porpoises. It is further conducive to the analysis of the finless porpoise organisms and habitats, and provides data support for the protection of the finless porpoise. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0031] Figure 2 This is a schematic diagram of the structure of the bracket of the present invention being arranged on a drone;

[0032] Figure 3 It is a schematic diagram of the relative relationship between the geographic coordinate system and the image coordinate system where the drone of the present invention is located. DETAILED DESCRIPTION

[0033] The specific implementation of the present invention is described in detail below with reference to the accompanying drawings and preferred embodiments.

[0034] like Figure 1 As shown, the present invention provides a method for identifying the position of a finless porpoise based on a cruise monitoring image of a drone. First, the drone is improved by adding two position sensors, and detection data containing finless porpoises is selected to obtain the route trajectory curve of the drone in the horizontal direction and the height trajectory curve in the vertical direction by fitting. Then, according to the shooting time T of the cruise monitoring image of the drone, the horizontal position coordinates (X, Y) and height information Z of the drone at this time, as well as the yaw angle κ and pitch angle ω of the drone are obtained. At the same time, according to the relative position information of the two position sensors, the roll angle of the drone is calculated. Finally, combined with the position coordinates (m, n) of the finless porpoise in the UAV cruise monitoring image, the position information corresponding to the shooting time T in the geographic coordinate system is calculated and converted into longitude and latitude information, thereby completing the position identification of the finless porpoise.

[0035] The details are as follows:

[0036] Step 1: Drone Transformation

[0037] The drone is provided with a camera for capturing images of porpoise patrol monitoring in the water, and a bracket can be installed on the fuselage of the drone along the vertical direction of its axis. The bracket can be made of carbon fiber material and is in a cross shape as a whole. The vertical rod of the bracket is arranged along the axis of the drone, and a position sensor, such as a GNSS-RTK device, is arranged at each end of the cross rod, which are respectively recorded as RTK-L and RTK-R (two GNSS-RTKs on the left and right). Figure 2 As shown, the distance between the two is recorded as D, and the relative height difference is recorded as h.

[0038] The center of the bracket coincides with the center of the drone to keep the drone balanced. The drone generally has a load capacity of 3kg, and the total weight of the two GNSS-RTK devices, namely the high-precision locator and the carbon fiber bracket, needs to meet the load requirements.

[0039] Step 2: Calculate the drone’s position information

[0040] Considering that the positioning frequency of GNSS-RTK equipment is usually 1 second, the flight speed of the drone is usually 5m / s, and the camera can take about 30 photos in 1 second, the flight distance of the drone is 5m. Therefore, to calculate the position of the finless porpoise in any photo, it is necessary to calculate the drone position (X, Y), altitude information Z, yaw angle κ, pitch angle ω, roll angle at the time of the photo shooting

[0041] To this end, we selected cruise monitoring images of finless porpoises from the videos taken by the drone's camera, recorded the shooting time T, selected the detection data corresponding to multiple consecutive detection times of the two position sensors before and after the shooting time T, and used the average value of the horizontal position and the average value of the height information in the detection data as fitting data. Using the least squares method, we obtained the route trajectory curve and the altitude trajectory curve.

[0042] If T is an integer second, select the horizontal plane position coordinates and height information Z of the drone within 7 consecutive seconds of T-3 seconds, T-2 seconds, T-1 second, T seconds, T+1 second, T+2 seconds, and T+3 seconds; if T is not an integer second, it is rounded to TT, then select the horizontal plane position coordinates and height information Z of the drone within 7 consecutive seconds of TT-3 seconds, TT-1 second, TT seconds, TT+1 second, TT+2 seconds, and TT+3 seconds, and then use the horizontal plane position coordinates of these 7 consecutive drones and the least squares fitting of the B-spline curve to obtain the route trajectory curve of the drone in this time period, and finally substitute the shooting time T into the route trajectory curve to calculate the horizontal position coordinates (X, Y) of the drone at this time;

[0043] Similarly, using the altitude information Z of these seven consecutive drones and the least squares fitting of the B-spline curve, the altitude trajectory curve of the drone during this period of time is obtained. Finally, the shooting time T is substituted into the altitude trajectory curve to calculate the altitude information Z of the drone at this time.

[0044] Step 3: Calculate the attitude angle of the drone: yaw angle κ, pitch angle ω, roll angle

[0045] The angle between the tangent direction of the trajectory curve at the shooting time T and the axis direction of the aircraft is taken as the yaw angle κ of the UAV;

[0046] The angle between the tangent direction of the height trajectory curve at the shooting time T and the horizontal direction is taken as the pitch angle ω of the UAV;

[0047] Calculate the roll angle of the drone based on the distance D and relative height difference h between the two position sensors

[0048] Step 4: Identify the location of the finless porpoise

[0049] The image coordinate system is established with the center point of the cruise monitoring image as the coordinate origin, the forward direction of the drone passing through the coordinate origin as the y-axis, and the right direction pointing to the flight direction through the coordinate origin as the x-axis, as shown in the following figure: Figure 3 As shown, the position coordinates (m, n) of the finless porpoise in the image coordinate system, that is, the pixel coordinates (m, n), are calculated.

[0050] The following equation is used to calculate the position information (X j ,Y j ).

[0051]

[0052] b1=cosωsinκ

[0053] b2=cosωcosκ

[0054] b3=-sinω

[0055]

[0056] The finless porpoise position coordinates calculated above are a plane coordinate system coordinates, which belong to the same plane coordinate system as the position of the drone. Both are geographic location coordinate systems, but longitude and latitude coordinates are often used in monitoring. Therefore, the following equation can be used to convert the finless porpoise position information (X j ,Y j ) to latitude and longitude (B,L)

[0057] The equation for coordinate transformation is as follows:

[0058]

[0059] Among them, B needs to be calculated iteratively, N is the radius of curvature of the earth's angular momentum, and e is the first eccentricity of the earth, which can all be obtained by looking up the table.

[0060] Iterative calculation: When using the above formula to calculate the geodetic latitude B, an iterative method is usually used. Use the initial value B1 of B to calculate N1 and sinB1, and then perform the second iteration according to the above formula until the difference between the last two B values ​​is less than the allowable value.

[0061] Although specific embodiments of the present invention are described above, those skilled in the art should understand that these are merely examples and that various changes or modifications may be made to these embodiments without departing from the principles and essence of the present invention. Therefore, the scope of protection of the present invention is limited by the appended claims.

Claims

1. A method for identifying the position of finless porpoises based on drone cruise monitoring images, characterized in that The following steps are involved: Step 1: Two position sensors are arranged in the vertical direction of the axis of the drone; Step 2: According to the detection data of the two position sensors, the horizontal route trajectory curve and the vertical height trajectory curve of the UAV are fitted, and then according to the shooting time T of the UAV cruise monitoring image, the horizontal position coordinates (X, Y) and height information Z of the UAV at this time are obtained; Step 3: Calculate the yaw angle k of the drone based on the route trajectory curve, calculate the pitch angle ω of the drone based on the altitude trajectory curve, and calculate the roll angle of the drone based on the relative position information of the two position sensors. Step 4: Calculate the position coordinates (m, n) of the finless porpoise in the drone cruise monitoring image, and then combine the yaw angle k, pitch angle ω and roll angle of the drone The location information of the finless porpoise corresponding to the shooting time T in the geographic coordinate system is calculated and converted into longitude and latitude information to complete the location identification of the finless porpoise.

2. The method for identifying the position of finless porpoises based on drone cruise monitoring images according to claim 1 is characterized in that: In the step 3, the angle between the tangent direction of the route trajectory curve at the shooting time T and the axis direction of the aircraft is taken as the yaw angle k of the drone; The angle between the tangent direction of the height trajectory curve at the shooting time T and the horizontal direction is taken as the pitch angle ω of the UAV; Calculate the roll angle of the drone based on the distance D and relative height difference h between the two position sensors 3. The method for identifying the position of finless porpoises based on drone cruise monitoring images according to claim 2 is characterized by: In the step 2, detection data corresponding to two position sensors at multiple consecutive detection moments are selected, and the cruise monitoring image corresponding to the detection data contains finless porpoises. The average value of the horizontal position and the average value of the height information in the detection data are used as fitting data, and the least squares method is used to obtain the route trajectory curve and the height trajectory curve.

4. The method for identifying the position of finless porpoises based on drone cruise monitoring images according to claim 2 is characterized in that: The center point of the cruise monitoring image is taken as the coordinate origin, the forward direction of the UAV passing through the coordinate origin is taken as the y-axis, and the right direction pointing to the flight direction through the coordinate origin is taken as the x-axis. The image coordinate system is established, and the position coordinates (m, n) of the finless porpoise in the image coordinate system are calculated. The following equation is used to calculate the position information (X j ,Y j ). b1=cosωsinκ b2=cosωcosκ b3=-sinω 5. The method for identifying the position of finless porpoises based on drone cruise monitoring images according to claim 4 is characterized in that: The position information of the finless porpoise (X j ,Y j ) to latitude and longitude (B,L) Among them, B needs to be calculated iteratively, N is the radius of curvature of the earth's angular circle, and e is the first eccentricity of the earth.

6. The method for identifying the position of finless porpoises based on drone cruise monitoring images according to claim 1 is characterized in that: In step one, two position sensors are arranged on the drone using a cross-shaped bracket, wherein the vertical rod of the cross-shaped bracket is arranged along the axis of the drone, and the two position sensors are arranged at the ends of the horizontal rod of the cross-shaped bracket.

7. The method for identifying the position of finless porpoises based on drone cruise monitoring images according to claim 6 is characterized by: The position sensor adopts an RTK locator, and the drone is provided with a camera for taking images of the patrol monitoring of the finless porpoise in the water.

Images