GPU-based integrated processing method and system for geochemical data

Patent metadata

Field	Value
Publication number	US-12307569-B2
Application number	US-202418596686-A
Country	US
Kind code	B2
Filing date	Mar 6, 2024
Priority date	Mar 8, 2023
Publication date	May 20, 2025
Grant date	May 20, 2025

Abstract

A GPU-based integrated processing method for geochemical data, includes the following steps: S 1 , storing geochemical data in a format of a preset three-dimensional array data structure Struct_A, S 2 , iteratively processing the formatted geochemical data by using a compute shader in a GPU, calculating a mean value and a standard deviation of content of each chemical element, calculating a threshold value of the chemical element, eliminating sampling points according to the threshold value, repeating the operations until there is no sampling point to be eliminated, and calculating a final mean value as a background value and a final threshold value as a lower limit value of anomaly; and S 3 , constructing spatial coordinates according to the obtained new geochemical data set, loading the spatial coordinates into a fragment shader for rendering, and completing anomaly delineation according to the background value and the lower limit value of anomaly.

First claim

What is claimed is:

1. A graphics processing unit (GPU)-based integrated processing method for geochemical data, comprising:
S 1 : storing geochemical data in a format of a preset three-dimensional array data structure Struct_A, wherein a first dimension of the three-dimensional array data structure Struct_A represents sampling points, a second dimension represents chemical elements, and a third dimension is used for counting sampling points and storing chemical element values;
S 2 : iteratively processing the formatted geochemical data by using a compute shader in a GPU, to be specific, performing pairwise parallel iterative addition on adjacent sampling points until an iterative sum of content of each chemical element in all sampling points is calculated, accordingly calculating a mean value and a standard deviation of content of each chemical element, calculating a threshold value of the chemical element, eliminating sampling points having a chemical element content below the threshold value, repeating the operations until there is no sampling point to be eliminated to obtain a new geochemical data set, and calculating a final mean value of the new geochemical data set as a background value and a final threshold value of the new geochemical data set as a lower limit value of anomaly; and
S 3 : constructing spatial coordinates according to the new geochemical data set, loading the spatial coordinates into a fragment shader for rendering, and completing anomaly delineation according to the background value and the lower limit value of anomaly;
wherein the step S 3 specifically comprises:
S 31 : creating a vertex buffer object (VBO) based on the new geochemical data set, wherein the spatial coordinates are used as vertex coordinates and the chemical element values are used as vertex attributes;
S 32 : triangulating the new geochemical data set, generating a list of vertex numbers for triangle drawing, and performing data interpolating in the process to obtain a smooth anomaly map to be delineated; and
S 33 : according to the background value and the lower limit value of anomaly, coloring the anomaly map to be delineated in the fragment shader; when an attribute value of a vertex is greater than the background value or the lower limit value of anomaly, setting a pixel of the vertex as an anomalous color; otherwise, setting the pixel as a non-anomalous color, to complete the anomaly delineation.

2. The GPU-based integrated processing method for geochemical data according to claim 1 , wherein the step S 2 specifically comprises:
S 21 : creating a first buffer and a second buffer accessible to the compute shader, writing an initial geochemical data set CSeti in the format Struct_A into the first buffer, performing pairwise parallel vector summation on sampling point count values C and chemical element values V of adjacent sampling points, storing results in the second buffer, performing pairwise summation on data in the second buffer and storing results in the first buffer, repeating the operations until a vector sum of the sampling point count values C and the chemical element values V of all the chemical elements in all the sampling points is calculated, and calculating a mean value PM of each chemical element by determining a ratio of the C to the V;
S 22 : creating a third buffer and a fourth buffer accessible to the compute shader, writing the initial geochemical data set CSet 1 into the third buffer, parallelly calculating a sum of squared differences between the chemical element values V of each chemical element and the corresponding mean value PM in adjacent sampling points and calculating a vector sum of the sampling point count values C in adjacent sampling points, storing results in the fourth buffer, performing pairwise parallel summation on adjacent sampling points in the fourth buffer and storing results in the third buffer, repeating the operations until a sum of squared differences between the chemical element values V and the mean value PM and a vector sum of the sampling point count values C in all sampling points are calculated, and accordingly calculating a standard deviation PSd of each chemical element; calculating the threshold value according to the mean value PM and the standard deviation PSd, and filtering the initial geochemical data set CSet 1 according to the threshold value to obtain a filtered geochemical data set CSet 2 ; and
S 23 : filtering the filtered geochemical data set CSet 2 in a same manner until there is no sampling point to be eliminated; and obtaining the final mean value PMN of each chemical element as the background value, and the final threshold value PtN of each chemical element as the lower limit value of anomaly.

3. The GPU-based integrated processing method for geochemical data according to claim 2 , wherein the sampling point count values C of all chemical elements in the initial geochemical data set CSet 1 are initialized to 1; and the sampling point count values C and the chemical element values V of all chemical elements in the second buffer and the fourth buffer are initialized to 0.

4. The GPU-based integrated processing method for geochemical data according to claim 2 , wherein a memory offset of the first buffer is twice of a memory offset of the second buffer.

5. The GPU-based integrated processing method for geochemical data according to claim 2 , wherein a memory offset of the third buffer is twice of a memory offset of the fourth buffer.

6. The GPU-based integrated processing method for geochemical data according to claim 2 , wherein when a chemical element value of a chemical element in adjacent sampling points is 0, a sampling point count value of the chemical element is also set to 0.

7. The GPU-based integrated processing method for geochemical data according to claim 1 , wherein when the anomaly map to be delineated is a multi-chemical element anomaly map, multiple color mixing methods are selected according to actual situations.

8. A GPU-based integrated processing apparatus for geochemical data, wherein the apparatus comprises:
a three-dimensional array module configured to store geochemical data in a format of a preset three-dimensional array data structure Struct_A, wherein a first dimension of the three-dimensional array data structure Struct_A represents sampling points, a second dimension represents chemical elements, and a third dimension is used for counting sampling points and storing chemical element values;
a background value calculation module configured to iteratively process the formatted geochemical data by using a compute shader in a GPU, to be specific, perform pairwise parallel iterative addition on adjacent sampling points until an iterative sum of content of each chemical element in all sampling points is calculated, accordingly calculate a mean value and a standard deviation of content of each chemical element, calculate a threshold value of the chemical element, eliminate sampling points having a chemical element content below the threshold value, repeat the operations until there is no sampling point to be eliminated to obtain a new geochemical data set, and calculate a final mean value of the new geochemical data set as a background value and a final threshold value of the new geochemical data set as a lower limit value of anomaly; and
an anomaly delineation module configured to construct spatial coordinates according to the new geochemical data set, load the spatial coordinates into a fragment shader for rendering, and complete anomaly delineation according to the background value and the lower limit value of anomaly; the anomaly delineation module has been specifically used to creating a vertex buffer object (VBO) based on the new geochemical data set, wherein the spati

Assignees

• Univ China Geosciences Beijing
• Tibet Julong Copper Co Ltd

Inventors

• Zheng Youye
• Xue Zhaolong
• Luosang Jiancuo
• Wei Jiangang
• Li Jingjing
• Kang Peng
• Suolang Zhuoga

Classifications

• G06T2219/2012

  Colour editing, changing, or manipulating; Use of colour codes · CPC title

• G06T19/20

  Editing of three-dimensional [3D] images, e.g. changing shapes or colours, aligning objects or positioning parts · CPC title

• G06T7/90

  Determination of colour characteristics · CPC title

• G06T15/87

  Gouraud shading · CPC title

• G06T15/205

  Image-based rendering · CPC title