Search

CN-122023502-A - Enteromorpha area estimation method based on unmanned aerial vehicle non-orthographic operation mode

CN122023502ACN 122023502 ACN122023502 ACN 122023502ACN-122023502-A

Abstract

The application relates to the technical field of marine ecological disaster monitoring, and discloses an enteromorpha area estimation method based on an unmanned aerial vehicle non-orthographic operation mode, which comprises the steps of firstly, judging a sea state mode by utilizing spectral energy analysis and inverting a basic wave height, and constructing a vertical non-uniform atmospheric extinction model for physical defogging; and finally, integrating the pixel-level macroscopic oblique projection, the wave shielding compensation factor and the microscopic surface area correction factor, and calculating the physical total surface area of the enteromorpha by combining a Sigmoid smooth cutoff weight function. The application solves the metering error caused by non-uniform atmospheric medium, wave shielding and microscopic geometric characteristic deletion under non-orthographic observation, and remarkably improves the monitoring precision under complex sea conditions.

Inventors

  • TAN WEI

Assignees

  • 青岛远度智能科技有限公司

Dates

Publication Date
20260512
Application Date
20260409

Claims (10)

  1. 1. An enteromorpha area estimation method based on an unmanned aerial vehicle non-orthographic operation mode is characterized by comprising the following steps of: Carrying out frequency domain analysis on an original non-orthographic image acquired by a monocular camera mounted on an unmanned aerial vehicle flight platform by utilizing a spectral energy grading locking module, extracting spectral energy density of a long-wave surge frequency band, comparing the spectral energy density with a preset steady sea state energy threshold to judge a sea state mode, and calculating or setting a basic wave height parameter according to the sea state mode; Constructing a vertical non-uniform atmospheric extinction coefficient distribution model according to the basic wave height parameter by utilizing a sea-air interface coupling depth field reconstruction module, carrying out non-uniform integration on the atmospheric extinction coefficient distribution model along a sight path to calculate scene atmospheric transmittance, and carrying out physical defogging treatment on an original non-orthographic image by utilizing the scene atmospheric transmittance to generate a restored image; Extracting texture entropy features and flare statistical features of the restored image by utilizing a fluid-solid coupling parameter inversion module, inverting a fluid-solid coupling damping coefficient of enteromorpha on waves based on the texture entropy features, and correcting microscopic equivalent mean square gradient of the enteromorpha surface by combining the flare statistical features; And comprehensively calculating the macroscopic projection area, the wave shielding compensation factor and the microscopic surface area correction factor of the pixel level by utilizing a multi-factor weighted area integration module, and carrying out weighted accumulation calculation on the macroscopic projection area, the wave shielding compensation factor, the microscopic surface area correction factor and the weight value calculated by the smooth cutoff weight function by combining a smooth cutoff weight function based on a Sigmoid function to output the enteromorpha physical total surface area.
  2. 2. The method for estimating the enteromorpha area based on the unmanned aerial vehicle non-orthographic operation mode according to claim 1, wherein the step of performing frequency domain analysis on the original non-orthographic image acquired by the monocular camera by using the spectral energy hierarchical locking module comprises the following steps: Performing two-dimensional discrete Fourier transform on the original non-orthographic image to obtain a frequency domain image; constructing an adaptive low-pass filter, wherein the cut-off frequency of the adaptive low-pass filter is set according to the projection pixel period of the minimum physical wavelength of the target surge on the image plane; Filtering the frequency domain image by using the self-adaptive low-pass filter, and extracting a low-frequency component corresponding to a long-wave surge frequency band; calculating an amplitude mode square mean value of the low-frequency component as a frequency spectrum energy density; if the frequency spectrum energy density is larger than the steady sea state energy threshold value, judging that the wind wave mode is adopted; and if the frequency spectrum energy density is smaller than or equal to the steady sea state energy threshold value, judging the mode as a steady mode.
  3. 3. The method for estimating the area of the enteromorpha based on the unmanned aerial vehicle non-orthographic operation mode according to claim 2, wherein the step of calculating or setting the basic wave height parameter according to the sea state mode comprises the following steps: when the wind wave mode is determined, carrying out inverse transformation on the low-frequency component to restore a spatial domain surge image, detecting the pixel spacing of adjacent wave crests on the spatial domain surge image, inverting the pixel spacing into physical wavelength by utilizing a perspective imaging geometrical relationship and a camera downward-looking inclination angle, and calculating the basic wave height parameter according to a preset wave steep coefficient; and when the stationary mode is determined, the basic wave height parameter is directly set to a preset tiny value, so that the subsequent atmospheric extinction coefficient distribution model is retracted to a standard state only containing background aerosol.
  4. 4. The method for estimating the area of the enteromorpha based on the unmanned aerial vehicle non-orthographic operation mode according to claim 1, wherein the step of constructing the vertical non-uniform atmospheric extinction coefficient distribution model according to the basic wave height parameter comprises the following steps: Establishing an extinction function driven by a vertical height independent variable, wherein the extinction function is formed by overlapping a background atmosphere extinction term and a droplet aerosol extinction term; The method comprises the steps of establishing a mapping relation between a sight line path and a vertical height according to a sight line incidence angle of a pixel point and an oblique Euclidean distance corresponding to the pixel point, and carrying out line integration on a extinction function along the sight line path to obtain scene atmosphere transmittance changing along the oblique Euclidean distance.
  5. 5. The method for estimating the area of the enteromorpha based on the unmanned aerial vehicle non-orthographic operation mode according to claim 1, wherein the step of inverting the fluid-solid coupling damping coefficient of the enteromorpha to waves based on the texture entropy characteristics comprises the following steps: Dividing the restored image into an enteromorpha target area and a seawater background area, and respectively calculating the local texture entropy of the enteromorpha target area and the local texture entropy of the seawater background area; constructing a mapping relation between the fluid-solid coupling damping coefficient and the texture entropy difference ratio; If the local texture entropy of the enteromorpha target area is obviously smaller than that of the sea water background area, judging that the fluid-solid coupling damping coefficient approaches to a complete damping state; if the local texture entropy of the enteromorpha target area is close to the local texture entropy of the sea water background area, judging that the fluid-solid coupling damping coefficient is approaching to an undamped state.
  6. 6. The method for estimating the area of the enteromorpha based on the unmanned aerial vehicle non-orthographic operation mode according to claim 5, wherein the step of correcting the microscopic equivalent mean square gradient of the surface of the enteromorpha by combining the flare statistics features comprises: Calculating the brightness distribution variance of a flare region in the restored image, and inverting the full spectrum mean square gradient of the seawater by using a Cox-Munk model; Carrying out attenuation correction on the full-spectrum mean square gradient of the seawater by utilizing the fluid-solid coupling damping coefficient to obtain the microscopic equivalent mean square gradient of the enteromorpha surface; According to the random surface geometry theory, calculating a microscopic surface area correction factor by utilizing the microscopic equivalent mean square gradient, wherein the microscopic surface area correction factor is used for representing the area increment of microscopic roughness relative to a horizontal projection plane.
  7. 7. The method for estimating an area of Enteromorpha prolifera based on a non-orthographic operation mode of an unmanned aerial vehicle according to claim 1, wherein the step of calculating a macroscopic projected area of a pixel level comprises: Determining an oblique Euclidean distance from a pixel point to a camera optical center and a sight line incident angle based on a pinhole camera imaging model; Calculating the ratio of the square of the oblique Euclidean distance to the normalized focal length of the camera, and dividing the ratio by the cosine value of the incident angle of the sight to obtain the macroscopic projection area; wherein the cosine of the angle of incidence of the line of sight is used to transform a projection of the cross-sectional area perpendicular to the line of sight to a horizontal sea surface.
  8. 8. The method for estimating the area of Enteromorpha prolifera based on the unmanned aerial vehicle non-orthographic operation mode according to claim 7, wherein the step of calculating the wave shielding compensation factor comprises the following steps: Constructing a shielding model based on probability statistics, wherein the wave shielding compensation factor is positively correlated with a basic wave height parameter, is positively correlated with a tangent value of a sight line incident angle corresponding to the macroscopic projection area, and is inversely proportional to the diagonal Euclidean distance of a pixel point; and carrying out numerical compensation on the enteromorpha area which is not captured by the camera due to wave crest shielding by utilizing the wave shielding compensation factor.
  9. 9. The method for estimating the area of the enteromorpha based on the unmanned aerial vehicle non-orthographic operation mode according to claim 7, wherein the step of combining the smooth truncated weight function based on the Sigmoid function comprises the following steps: Defining an effective observation distance cut-off threshold and a transition bandwidth parameter; constructing a Sigmoid type weight function taking the diagonal Euclidean distance as an independent variable; The Sigmoid weight function approaches to the reserved original value in a short-distance area with the inclined Euclidean distance far smaller than the effective observation distance cut-off threshold value, approaches to the inhibition to zero in a long-distance area with the inclined Euclidean distance far larger than the effective observation distance cut-off threshold value, and smoothly descends in a transition area near the effective observation distance cut-off threshold value; and carrying out weighted inhibition on the area calculation result of the far-end low signal-to-noise ratio pixel by using the smooth truncation weight function so as to prevent the numerical value from diverging.
  10. 10. The enteromorpha area estimation method based on the unmanned aerial vehicle non-orthographic operation mode according to claim 1 is characterized in that the output enteromorpha physical total surface area is achieved by multiplying a macroscopic projection area, a wave shielding compensation factor, a microscopic surface area correction factor and weight values calculated according to a smooth truncated weight function of each pixel point identified as enteromorpha in the restored image by four terms to obtain a physical surface area contribution value of the pixel point, and accumulating the physical surface area contribution values of all the enteromorpha pixel points to obtain the enteromorpha physical total surface area.

Description

Enteromorpha area estimation method based on unmanned aerial vehicle non-orthographic operation mode Technical Field The invention relates to the technical field of marine ecological disaster monitoring, in particular to an enteromorpha area estimation method based on an unmanned aerial vehicle non-orthographic operation mode. Background The unmanned aerial vehicle aerial remote sensing is an important means for monitoring the offshore enteromorpha disaster in a normalized manner, and has the advantages of high timeliness, flexibility, high resolution and the like compared with shore-based monitoring and satellite remote sensing. In practical operation, in order to improve the monitoring coverage width and efficiency of a single flight, the unmanned aerial vehicle generally adopts a wide-angle oblique photography (non-orthographic operation) mode. However, in such unmanned aerial vehicle non-orthographic observation scenes, existing image processing and area metering methods are difficult to adapt to the complex physical environment of the sea-air interface. Due to the fact that the near-shore sea area is often accompanied with the phenomenon of wind wave breaking, salt mist and spray aerosols which are remarkably uneven in vertical distribution exist in the near-surface atmosphere, the traditional defogging algorithm is mostly based on the idealized assumption of the uniform distribution of atmospheric media, so that the existing deviation is calculated through the transmittance on the unmanned plane oblique path observation path, and further accurate inversion of scene depth information is seriously affected. In addition, under the low-angle oblique observation condition, dynamic fluctuation of sea surface waves can generate a shielding effect on a target floating in a trough, the sea surface is simplified into a static rigid plane to carry out homography geometric projection in the prior art, a view field blind area caused by wave shielding is not considered, and the area increment caused by the surface roughness of enteromorpha serving as a viscoelastic floating body on a microscopic scale is ignored, so that systematic errors exist in finally calculated monitoring data, and the quantitative requirements of accurate disaster prevention and reduction are difficult to meet. Therefore, the invention provides an enteromorpha area estimation method based on a non-orthographic unmanned aerial vehicle operation mode, which solves the defects in the prior art. Disclosure of Invention Aiming at the defects of the prior art, the invention provides an enteromorpha area estimation method based on an unmanned aerial vehicle non-orthographic operation mode, which solves the problem of low enteromorpha area metering precision caused by non-uniform distribution of an atmospheric medium, dynamic wave shielding and missing of geometric features of a microscopic surface in an unmanned aerial vehicle non-orthographic observation scene. In order to achieve the aim, the invention is realized through the following technical scheme that the enteromorpha area estimation method based on the unmanned aerial vehicle non-orthographic operation mode comprises the following steps: Carrying out frequency domain analysis on an original non-orthographic image acquired by a monocular camera by utilizing a spectral energy grading locking module, extracting spectral energy density of a long-wave surge frequency band, comparing the spectral energy density with a preset steady sea state energy threshold to judge a sea state mode, and calculating or setting a basic wave height parameter according to the sea state mode; Constructing a vertical non-uniform atmospheric extinction coefficient distribution model according to the basic wave height parameter by utilizing a sea-air interface coupling depth field reconstruction module, carrying out non-uniform integration on the atmospheric extinction coefficient distribution model along a sight path to calculate scene atmospheric transmittance, and carrying out physical defogging treatment on an original non-orthographic image by utilizing the scene atmospheric transmittance to generate a restored image; Extracting texture entropy features and flare statistical features of the restored image by utilizing a fluid-solid coupling parameter inversion module, inverting a fluid-solid coupling damping coefficient of enteromorpha on waves based on the texture entropy features, and correcting microscopic equivalent mean square gradient of the enteromorpha surface by combining the flare statistical features; And comprehensively calculating the macroscopic projection area, the wave shielding compensation factor and the microscopic surface area correction factor of the pixel level by utilizing a multi-factor weighted area integration module, and carrying out weighted accumulation calculation on the macroscopic projection area, the wave shielding compensation factor, the microscopic surface area correction factor