Search

CN-121784701-B - Self-adaptive deblurring method, device and storage medium for radial velocity of isolated echo

CN121784701BCN 121784701 BCN121784701 BCN 121784701BCN-121784701-B

Abstract

The invention discloses a self-adaptive deblurring method, self-adaptive deblurring equipment and a storage medium for radial speed of an isolated echo, and relates to the technical field of radar signal processing. The method comprises the steps of firstly identifying a main echo region and an isolated echo region from a preprocessed radar radial velocity matrix, establishing a trusted region set, dividing the isolated echo region into two types according to velocity distribution characteristics of the isolated echo region, respectively processing the isolated echo region, predicting initial velocity by utilizing local radial data, guiding to conduct azimuth propagation type pixel-by-pixel deblurring, and then bringing the radar region into the trusted region set after completion, determining anchoring velocity according to spatial relation between the radar region and adjacent pixels in the trusted region set for a second type region, and carrying out integral symbol consistency correction. The method realizes the automatic and accurate de-blurring of various isolated echoes, solves the problem of failure of the traditional method in an isolated target scene, and remarkably improves the quality of radar velocity fields.

Inventors

  • LIU SIRUI
  • CUN HUAICHENG
  • Hu Henglin

Assignees

  • 浙江华盛雷达股份有限公司

Dates

Publication Date
20260512
Application Date
20260306

Claims (10)

  1. 1. An adaptive deblurring method for isolated echo radial velocity, the method comprising: Acquiring a radar radial velocity matrix subjected to the pre-back blurring processing, and identifying a main echo region and at least one isolated echo region based on the radial velocity matrix; initializing a trusted region set, and marking all pixels of the main echo region as trusted; Dividing the isolated echo regions into a first type region and a second type region according to the statistical distribution characteristics of the velocity values in the isolated echo regions; the first deblurring processing is carried out on the first type of areas, namely, for each first type of area, an initial speed is predicted based on local radial data of the first type of area, the initial speed is taken as a guide, and the adjacent radial propagation correction operation covered by the area is carried out along an azimuth angle to finish the pixel-by-pixel deblurring of the speed value in the area; And executing second deblurring processing on the second type of regions, namely determining an anchoring speed according to the spatial adjacent relation between each second type of region and the trusted pixels in the trusted region set based on the updated trusted region set, and performing symbol consistency correction on the speed values in the region according to the symbol relation of the integral speed statistic value of the region relative to the anchoring speed.
  2. 2. The adaptive de-blurring method of isolated echo radial velocity of claim 1 wherein identifying a main echo region and at least one isolated echo region based on the radial velocity matrix comprises: Generating a binary effective mask matrix of the radial velocity matrix; the radial velocity matrix and the effective mask matrix are respectively subjected to the same azimuth period expansion to generate a corresponding expansion velocity matrix and an expansion mask matrix, wherein the azimuth period expansion is realized by respectively adding one row before the first row and after the last row of the original matrix and respectively copying the last row and the first row of the original matrix to the newly added first row and last row; And carrying out connected domain analysis based on the extended mask matrix, identifying all connected domains, determining the connected domain with the largest area as a main echo region, and determining the rest connected domains as isolated echo regions.
  3. 3. The adaptive de-blurring method of isolated echo radial velocity of claim 1 wherein predicting a starting velocity based on local radial data of a region of a first type comprises: Determining one radial direction which contains most pixels in the first type of region as a central radial direction; in the central radial direction, determining a pixel which belongs to the first type area and is nearest to the radar end, and taking a column index of the pixel as an initial distance gate; Searching a section of continuous effective speed sequence from the near radar end of the initial range gate in the center radial direction; Based on the effective speed sequence, the predicted speed at the starting distance gate is obtained through fitting or statistical extrapolation, and the predicted speed is taken as the starting speed.
  4. 4. The adaptive de-blurring method of isolated echo radial velocity of claim 3 wherein searching for a continuous sequence of valid velocities using a multi-stage degradation strategy comprises: Searching a first speed sequence meeting a first quality condition, wherein the first quality condition comprises a sequence which is continuous, the speed standard deviation of the sequence is lower than a first standard deviation threshold value and the length of the sequence is not smaller than a length threshold value; If the first speed sequence fails to search, searching a second speed sequence meeting a second quality condition, wherein the second quality condition comprises continuous sequences, the speed standard deviation of the second speed sequence is lower than a second standard deviation threshold value, the length of the sequence is not smaller than the length threshold value, and the second standard deviation threshold value is larger than a first standard deviation threshold value; and collecting the first-searched continuous effective speed values with the number equal to the length threshold value, and taking the median of the collected speed values as the predicted speed.
  5. 5. The adaptive de-blurring method of isolated echo radial velocity of claim 1 wherein the operation of correcting adjacent radial propagation in azimuth to the area of coverage comprises: Searching a pixel speed meeting a similarity condition and an isotropy condition as a reference speed based on the initial speed in the center radial direction; the method comprises the steps of sequentially correcting all pixels in a central radial direction according to a pixel-by-pixel correction rule to obtain a reference radial direction, wherein the pixel-by-pixel correction rule is that for a first pixel, the absolute difference between the original speed and the reference speed is calculated, the correction amount is determined according to the comparison result of the absolute difference and the Nyquist speed of a preset multiple, and the reference speed is updated to be the corrected speed of the pixel after correction; And for any pixel in the radial direction to be corrected, dynamically selecting at least one nearest neighbor reference speed from corrected pixels in the spatial neighborhood of the pixel, and correcting by applying the pixel-by-pixel correction rule.
  6. 6. The adaptive deblurring method of isolated echo radial velocity according to any one of claims 1 to 5, wherein performing a first deblurring process on a first type of region further comprises performing a discrete point correction step after the pixel-by-pixel deblurring is completed: calculating the median of the corrected speeds of all pixels in the first type region; And traversing all pixels in the first type region, and if the corrected speed of a certain pixel is opposite to the sign of the median, applying one-time sign inversion correction to the pixel to enable the sign of the corrected speed of the pixel to be consistent with the sign of the median.
  7. 7. The adaptive de-blurring method of isolated echo radial velocity of claim 1 wherein determining an anchor velocity based on its spatial proximity to a trusted pixel in the set of trusted regions comprises: calculating the geometric center coordinates of the second type of region; searching all trusted pixels with the space distance smaller than a preset distance threshold value from the geometric center in the trusted region set by taking the geometric center as a reference; the median of the velocity values of all the trusted pixels found is calculated and is determined as the anchoring velocity.
  8. 8. The adaptive de-blurring method of isolated echo radial velocity of claim 1 wherein performing a sign consistency correction on velocity values within the region based on a sign relationship of an overall velocity statistic of the region relative to the anchor velocity comprises: calculating the average value of the original speeds of all pixels in the second type area; comparing the mean value to a sign of the anchor speed; and if the average value is opposite to the sign of the anchoring speed, applying one-time sign inversion correction to all pixels with the signs of the speeds opposite to the anchoring speed in the second type area, wherein the sign inversion correction is that a preset correction amount is added or subtracted on the original speed value of the pixel, so that the signs of the corrected speeds of the pixel are consistent with the signs of the anchoring speeds.
  9. 9. An electronic device comprising a memory, a processor and a computer program or instructions stored on the memory, wherein the processor executes the computer program or instructions to implement the adaptive de-blurring method of orphan echo radial velocity of any of claims 1-8.
  10. 10. A computer readable storage medium having stored thereon a computer program or instructions which, when executed by a processor, implements the adaptive deblurring method of isolated echo radial velocity according to any one of claims 1 to 8.

Description

Self-adaptive deblurring method, device and storage medium for radial velocity of isolated echo Technical Field The invention belongs to the technical field of radar signal processing, and particularly relates to a self-adaptive deblurring method, equipment and a storage medium for radial speed of an isolated echo. Background The pulse Doppler weather radar can acquire the reflectivity factor representing the rainfall intensity by transmitting and receiving electromagnetic waves, and can extract the radial speed revealing the motion state of the atmospheric flow field based on the Doppler effect. The parameter is the most direct and key physical quantity for three-dimensional wind field inversion, medium-scale vortex (such as tornado and mesocyclone), monitoring of wind shear, downburst and other disastrous weather phenomena. The quality of radar data, especially the accuracy of radial velocity, directly determines the timeliness and reliability of the short-term forecasting and early warning. However, limited to pulse repetition frequency, there is an insurmountable theoretical upper limit on the velocity measurement capability of doppler weather radar, namely nyquist velocity (±v N). When the true radial velocity of the precipitation particles or turbulence scatterers in the atmosphere is outside this range, the radar observations will jump periodically, a phenomenon known as "velocity ambiguity". Uncorrected fuzzy velocity fields distort the true airflow structure into unrecognizable chaotic images, so that all subsequent advanced products (such as wind field inversion and storm identification) lose scientific significance. Therefore, developing an efficient and robust deblurring algorithm is a core technical difficulty that must be overcome in the preprocessing link of radar data. To overcome this difficulty, a number of technological routes have been developed in the industry, but they all face serious challenges in dealing with the complexities of actual weather, especially the ubiquitous "orphan echoes": The core of the global reference method is a stable speed reference field provided by a large-range continuous strong echo. The method can be used for processing typhoons, lines and other large weather systems. However, it is a fatal disadvantage that once the back wave field is broken, and is distributed in isolation (e.g., local heat convection is primary, independent monomers in typhoon peripheral screw rain band, offshore decentralized precipitation), the algorithm fails completely by being unable to anchor a reliable global reference, even causing false global velocity reversals. The adjacent radial propagation method assumes that the wind field continuously changes between adjacent azimuth angles and propagates correction information through an "infectious" path. The success of which depends heavily on the continuity of the echo in azimuth. In actual observation, a topography obstruction, electromagnetic wave attenuation, or discontinuity of precipitation itself may cause a "fault" of echo data. Once the propagation chain is interrupted, the echo on the "island" becomes an information island, and no correction is obtained. This makes the method extremely vulnerable to processing power of isolated, weak echoes. And the external wind field constraint method is to carry out constraint by introducing external wind fields such as numerical forecasting or sounding and the like as priori knowledge. Although the method has a certain theoretical value, the method has fundamental limitations that firstly, external data and radar observation are seriously mismatched in space-time resolution, systematic deviation is easy to generate after the external data and the radar observation are introduced, and secondly, the real-time and autonomous processing capacity of the radar system is seriously damaged by relying on external data flow, so that the method has low operability in service application. In summary, the prior art commonly faces a prominent technical contradiction and bottleneck: In the face of small-scale, isolated and distributed discrete echo targets (which are key targets for strong convection monitoring), the existing algorithm generally has a 'deblurring blind zone' due to the lack of self-adaptive local reference establishment capability and intelligent region segmentation and cooperation mechanism. The method comprises the following steps of (1) enabling an algorithm to autonomously generate reliable local speed trend prediction for an isolated area without global reference, (2) enabling the processed reliable area to be incapable of being effectively identified and used for providing a dynamic anchor point for subsequent processing, (3) ignoring azimuth periodicity of radar scanning to cause artificial speed fault to an image edge, and (4) enabling residual speed island which is uncoordinated with a physical field after processing to lack of effective space consistency correction means. Ther