CN116739183B - A mine safety risk early warning and prediction system - Google Patents

Abstract

The invention provides a mine safety risk early warning and prediction system, which belongs to the technical field of safety early warning and includes: an image modeling module for image modeling of the mine mountain to obtain a preliminary mountain structure model; and a range determination module for determining the area. Correction range; the laser modeling module is used to model the regional correction range and obtain the regional correction range model; the orientation determination module is used to determine the fusion orientation of the regional correction range model in the preliminary mountain structure model; the area determination module is used to The deformation area of the mine is determined based on the complete mountain structure model; the hazard prediction module is used to predict the mine hazard area based on the current curvature and the corresponding initial curvature of the deformation area in the preset mountain structure model; the hazard warning module is used to predict the mine hazard area based on the mine hazard Send an early warning signal to the regional location; this invention is used to achieve more accurate early warning of areas where surface depression may be dangerous without being separated from the hardware in contact with the ground.

Description

A mine safety risk early warning and prediction system

Technical field

The invention relates to the technical field of safety early warning, and in particular to a mine safety risk early warning and prediction system.

Background technique

At present, with the development of industrialization and informatization, the demand for mineral resources is increasing, resulting in the continuous increase in the mining volume of mineral resources. The production capacity of many mining enterprises is constantly increasing. Since most mining enterprises already have a long service life Over the years, shallow mineral resources have been exhausted, and most of them are deep mineral resources. After the deep primary ore bodies and associated waste rocks are mined, underground spaces of varying sizes will be formed. Under the influence of gravity and the inconsistency of ground stress, Under the influence of equilibrium and other factors, ground cracks first occur in the goaf area, and gradually develop into ground collapse in the goaf area, which will lead to verified property safety accidents and even personal safety accidents. Therefore, mine safety risk early warning is crucial in the field of mining.

However, the existing mine safety risk early warning and prediction system only analyzes and provides early warning through various hardware in contact with the ground. Because the mine environment is harsh, the hardware in contact with the ground for a long time is easily corroded and damaged, making it difficult to provide accurate early warning. For example, the public number is " CN112885038A", published on June 1, 2021, is a Chinese invention patent with the patent name "Ground Collapse Early Warning Device and Ground Collapse Early Warning System". The method includes the following steps: a trigger unit, including a plurality of devices connected in sequence along the up and down direction. A trigger component, different trigger components corresponding to different strata; and an early warning unit, located on the top of the trigger unit, the early warning unit including a plurality of alarm components corresponding to the trigger component, each of the The alarm components are respectively connected to the corresponding trigger components through different connection components; when a trigger component falls off from the queue, it is pulled by the corresponding connection component under the action of gravity to trigger the corresponding alarm. component to determine the presence of a void in the formation corresponding to the triggering component. This invention can solve the problem that initial collapse in the lower part of the ground is not easily detected, and the gradual expansion of the collapse can easily cause major safety accidents. However, this patent cannot rely on a small number of equipment to conduct surface analysis of the collapse, and only provides analysis and early warning through many hardware in contact with the ground. Because the mining environment is harsh, hardware that has been in contact with the ground for a long time is easily rusted and damaged, so it is difficult to accurately provide early warning.

Therefore, the present invention proposes a mine safety risk early warning and prediction system, which is used to achieve more accurate early warning of areas where surface depression may be dangerous without being separated from the hardware in contact with the ground, so as to ensure the safe production of mines.

Contents of the invention

The invention provides a mine safety risk early warning and prediction system, which is used to perform preliminary modeling of the surface of the mine mountain through only image pairs of the mine mountain surface, and obtain a preliminary mountain structure model. The needs existing in the preliminary mountain structure model can be determined through the range determination module. For further remodeling, use the laser modeling module to model the area that needs to be remodeled to obtain a regional correction range model, and conduct more accurate modeling of areas where the model that appears during image modeling is difficult to reflect the real structure. This further ensures the accuracy and data integrity of the complete mountain structure model. The orientation determination module and the area determination module can be combined to determine the deformation area in the mine. The hazard prediction module is used to accurately predict the deformation area by analyzing it. In dangerous areas of mines, only relying on the surface data of mine mountains collected by drones can more accurately predict the areas where surface depressions are dangerous. That is, on the premise of being separated from the hardware in contact with the ground, it is possible to predict possible surface depressions. More accurate early warning of the area to ensure the safe production of mines.

The invention provides a mine safety risk early warning and prediction system, which includes:

The image modeling module is used to frequency decompose and fuse the plane-converted mine mountain images to obtain the fused image, and build a preliminary mountain structure model based on the fused image;

The range determination module is used to perform unit area decomposition and boundary enhancement on the fused image, obtain the boundary enhancement sub-image, and determine the regional correction range based on the correction value of the boundary enhancement sub-image;

The laser modeling module is used to re-model the regional correction range and obtain the regional correction range model;

The orientation determination module is used to process based on the preliminary mountain model and the regional correction range model, and determine the fusion orientation of the regional correction range model in the preliminary mountain structure model;

The area determination module is used to obtain a complete mountain structure model based on the fusion orientation, and determine the deformation area of the mine through the complete mountain structure model and the preset mountain structure model;

The hazard prediction module is used to calculate the depression curvature of the deformation area of the mine, obtain the current curvature of the deformation area, and predict the dangerous area of the mine based on the current curvature and the corresponding initial curvature of the deformation area in the preset mountain structure model;

Danger warning module, used to send early warning signals based on the location of dangerous areas in mines.

Preferably, a mine safety risk early warning and prediction system and image modeling module include:

The image acquisition submodule is used to collect images of the area where the mine is located based on the shooting equipment set up on the drone, and obtain multiple images of the mine mountain;

The image conversion sub-module is used to input multiple mine mountain images into the preset plane conversion model to obtain corresponding splicable images;

The graphics fusion sub-module is used to frequency decompose all splicable images and then fuse them to obtain a fused image, and build a model based on the fused image to obtain a preliminary mountain structure model.

Preferably, a mine safety risk early warning and prediction system, the graphics fusion submodule includes:

The image conversion unit is used to perform Fourier transform on the splicable image, obtain the spectrogram, and determine the brightness value of each pixel in the spectrogram as the frequency component representation value in the splicable image;

The area determination unit is used to determine the same frequency component representation value in different splicable images, and the area surrounded by pixels corresponding to all the same frequency component representation values in each two splicable images is regarded as corresponding to two Stitchable areas between stitchable images;

The image splicing unit is used to splice all splicable images in pairs based on all splicable areas to obtain a fused image;

The model building unit is used to extract three-dimensional point clouds from the fused image, conduct modeling based on the three-dimensional point cloud data, and obtain a preliminary mountain structure model.

Preferably, a mine safety risk early warning and prediction system, a synthesis unit, includes:

An image generation subunit, used to generate the first sub-image corresponding to each frequency component of the splicable image based on the spectrogram;

The image distinction subunit is used to determine all the first sub-images of the low-frequency component and all the second sub-images of the high-frequency component in the splicable area in the two splicable images to be fused;

The low-frequency processing subunit is used to determine the average gray value of all pixels at the same position in all the first sub-images, and assign the average gray value to the pixels at the corresponding positions in the blank image to obtain a low-frequency fusion image. ;

The high-frequency processing subunit is used to determine the maximum gray value of all pixels at the same position in all the second part sub-images, and assign the maximum gray value to the pixels at the corresponding positions in the blank image, Obtain high-frequency fusion images;

The image splicing subunit overlays and synthesizes the low-frequency fusion image and the high-frequency fusion image to obtain a new splicable area. Based on the new splicable area, the two splicable images to be fused are spliced to obtain a fused image.

Preferably, a mine safety risk early warning and prediction system and a scope determination module include:

The boundary optimization submodule is used to divide the fusion image that constructs the preliminary mountain structure model into multiple second sub-images of unit area, perform image boundary detection on all second sub-images, and determine the boundary parts of all second sub-images. Perform boundary enhancement on the boundary parts of all second sub-images to obtain a set of boundary-enhanced sub-images;

The correction value calculation submodule is used to determine the correction value of each boundary enhancement sub-image in the boundary enhancement sub-image set based on the following formula:

;

In the formula, is the correction value of the boundary enhancer image, a is the ordinate value of the lower boundary of the boundary enhancer image in the plane rectangular coordinate system, b is the ordinate value of the upper boundary of the boundary enhancer image in the plane rectangular coordinate system, c is the abscissa value of the left boundary of the boundary enhancer image in the plane rectangular coordinate system, d is the abscissa value of the right boundary of the boundary enhancer image in the plane rectangular coordinate system,/> For the coordinates in the boundary enhancer image/> The maximum value of the pixel difference between the pixel at and each pixel in the enhanced boundary in the corresponding boundary enhanced sub-image,/> For the coordinates in the boundary enhancer image/> The minimum value of the pixel difference between the pixel at and each pixel in the enhanced boundary in the corresponding boundary enhancement sub-image,/> Differentiate for the independent variable;

The range judgment submodule is used to regard the boundary enhancement sub-images whose correction values are greater than the rated correction value in the boundary enhancement sub-image set as areas to be corrected, and to regard the corresponding area ranges of all areas to be corrected in the preliminary mountain structure model as area corrections. scope.

Preferably, a mine safety risk early warning and prediction system and laser modeling module include:

The laser sampling sub-module determines multiple scanning points in the regional correction range, controls the drone to scan at the actual position corresponding to the scanning point in the mine mountain, and obtains three-dimensional point cloud data of the corresponding scanning point;

The coincidence degree calculation sub-module is used to calculate the coincidence degree of the three-dimensional point cloud data of all adjacent scanning points;

The supplementary scanning sub-module is used to insert new scanning points between adjacent scanning points whose coincidence degree is less than the minimum coincidence degree threshold, and to supplementally obtain the three-dimensional point cloud data of the new scanning points until all adjacent scans are newly determined. When the degree of coincidence between the three-dimensional point cloud data of the points is not less than the minimum coincidence degree threshold, the first three-dimensional point cloud data is obtained;

The point cloud deduplication submodule is used to randomly remove a group of three-dimensional point cloud data of adjacent scanning points whose coincidence degree exceeds the maximum coincidence degree threshold in the first three-dimensional point cloud data, and obtain the second three-dimensional point cloud data;

A registration point sub-module, used to determine the same three-dimensional point cloud in the three-dimensional point cloud data of adjacent scanning points in the second three-dimensional point cloud data as the registration point corresponding to the adjacent scanning point;

The point cloud synthesis submodule is used to synthesize the three-dimensional point cloud data corresponding to adjacent scanning points based on the registration points to obtain a complete point cloud of the regional correction range;

The model construction submodule is used to build a regional correction range model based on the complete point cloud of the regional correction range.

Preferably, a mine safety risk early warning and prediction system and an orientation determination module include:

The size unification submodule is used to unify the size of the preliminary mountain structure model and regional correction range model;

The position determination submodule is used to determine the fusion orientation of the regional correction range model in the preliminary mountain structure model.

Preferably, a mine safety risk early warning and prediction system and an area determination module include:

The fusion sub-module fuses the regional correction range model and the preliminary mountain structure model based on the fusion orientation to obtain a complete mountain structure model;

The deformation determination submodule is used to compare the complete mountain structure model with the preset mountain structure model to determine the deformation area.

Preferably, a mine safety risk early warning and prediction system and hazard prediction module include:

The initial curvature calculation submodule is used to calculate the depression curvature of the deformation area based on the preset mountain structure model to obtain the initial curvature of the deformation area;

The depression curvature submodule is used to calculate the depression curvature based on the structural changes of the complete mountain structure model in the deformation area to obtain the current curvature of the deformation area;

The prediction submodule is used to predict the deformation area as a dangerous area of the mine when the difference between the current curvature and the initial curvature is greater than the preset value.

Preferably, a mine safety risk early warning and prediction system and hazard early warning module include:

The signal generation submodule is used to generate corresponding early warning signals based on the location of the mine's dangerous areas;

The signal sending submodule is used to send early warning signals to the monitoring and early warning background.

The beneficial effects produced by the present invention compared with the existing technology are as follows: preliminary modeling of the surface of the mine mountain is carried out only through image pairs of the surface of the mine mountain, and a preliminary mountain structure model is obtained. Through the range determination module, the needs existing in the preliminary mountain structure model can be determined. For the re-modeled area, use the laser modeling module to model the area that needs to be re-modelled, obtain the area correction range model, and conduct more accurate modeling of areas where the model that appears during image modeling is difficult to reflect the real structure, and further It ensures the accuracy and data integrity of the complete mountain structure model. The orientation determination module and the area determination module can be combined to determine the deformation area in the mine. The hazard prediction module is used to accurately predict the mine through the analysis of the deformation area. For dangerous areas, only relying on the mine mountain surface data collected by drones can more accurately predict the areas where surface depressions are dangerous. That is, on the premise of being separated from the hardware in contact with the ground, it is possible to predict possible surface depressions. More accurate early warning in the area to ensure the safe production of mines.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof.

The technical solutions of the present invention will be described in further detail below through examples.

Detailed ways

Preferred embodiments of the present invention are described below. It should be understood that the preferred embodiments described here are only used to illustrate and explain the present invention and are not intended to limit the present invention.

Embodiment 1: The present invention provides a mine safety risk early warning and prediction system, including:

The image modeling module is used to frequency decompose and fuse the plane-converted mine mountain images to obtain the fused image, and build a preliminary mountain structure model based on the fused image;

The range determination module is used to perform unit area decomposition and boundary enhancement on the fused image, obtain the boundary enhancement sub-image, and determine the regional correction range based on the correction value of the boundary enhancement sub-image;

The laser modeling module is used to re-model the regional correction range and obtain the regional correction range model;

The orientation determination module is used to process the preliminary mountain model and the regional correction range model, and determine the fusion orientation of the regional correction range model in the preliminary mountain structure model;

The area determination module is used to obtain a complete mountain structure model based on the fusion orientation, and determine the deformation area of the mine through the complete mountain structure model and the preset mountain structure model;

The hazard prediction module is used to calculate the depression curvature of the deformation area of the mine, obtain the current curvature of the deformation area, and predict the dangerous area of the mine based on the current curvature and the corresponding initial curvature of the deformation area in the preset mountain structure model;

Danger warning module, used to send early warning signals based on the location of dangerous areas in mines.

In this embodiment, the image of the mine mountain is an image of the surface of the mine mountain taken by a drone.

In this embodiment, the mine mountain is a complete mountain structure including the internal channel structure and the outer surface structure of the mine.

In this embodiment, the preliminary mountain structure model is a model obtained based on preliminary modeling of the image of the mine mountain surface. The model may fail to display the structure of some parts of it (due to obstructions in the mine mountain image). Case.

In this embodiment, the area correction range is an area where the model is difficult to reflect the real structure when modeling based on the image of the mine mountain surface, such as an area blocked by trees.

In this embodiment, the regional correction range is re-modelled as the UAV laser scanning device emits a laser signal to the target position in the deformation area, and then collects the reflected laser signal, and uses the speed of light, the laser from the emission to The returned time and other information are used to measure the distance information of the target position in the deformation area, and then combined with IMU, odometer, GNSS and other information to calculate the three-dimensional coordinate information of the target position in the deformation area, and then realize modeling.

In this embodiment, the regional correction range model is a model of a region in the preliminary mountain structure model that is difficult to reflect the real structure.

In this embodiment, the processing of the preliminary mountain model and the regional correction range model is to unify the size of the preliminary mountain model and the regional correction range model (for example, both are processed to 750 pixels). 422 pixel size) to ensure that they can be directly spliced and merged.

In this embodiment, the fusion orientation is the spliced and merged position of the preliminary mountain model and the regional correction range model.

In this embodiment, the complete mountain structure model is a model spliced and merged with the preliminary mountain model and the regional correction range model, which reflects the information of the complete mine mountain surface.

In this embodiment, the preset mountain structure model is a mountain structure model obtained by modeling the mountain surface of the mine through laser modeling equipment in advance.

In this embodiment, the deformation area is the area where the curvature of the mine mountain structure changes.

In this embodiment, the concave curvature is calculated by analyzing the preset mountain structure model part corresponding to the position of the deformation region based on the second basic formula of the existing curved surface, so that the position of any point in the deformation region (set as point a) can be The curvature of is equal to the curvature of the tangent vector direction curve at its position, and then the mathematical relationship between the curvature of point a in the deformation area and Gaussian curvature is determined based on Euler's formula (the curvature of any tangent direction passing through surface point a satisfies Euler's formula). According to Gaussian The mathematical relationship between curvature (Gaussian curvature is a real symmetric matrix, then it can be decomposed into the product of an orthogonal matrix and a diagonal matrix) and the curvature of point a in the deformation area can be obtained by performing the above process on all points in the deformation area. Traverse the curvature of all points in the deformation area, and set the maximum curvature as the concave curvature of the deformation area.

In this embodiment, the curvature data is the initial curvature and the current curvature of the deformation region.

In this embodiment, the dangerous area of the mine is an area where the mountain surface will collapse or the structural stress does not meet the overall stability requirements of the mountain.

The beneficial effects of the above technology are as follows: preliminary modeling of the surface of the mine mountain is carried out only through the image pair of the surface of the mine mountain, and a preliminary mountain structure model is obtained. The range determination module can determine the areas in the preliminary mountain structure model that need further remodeling. Use the laser modeling module to model the areas that need to be re-modelled, obtain the area correction range model, and conduct more accurate modeling of areas where the model that appears during image modeling is difficult to reflect the real structure, further ensuring the completeness of the built The accuracy and data integrity of the mountain structure model. The combination of the orientation determination module and the area determination module can determine the deformation area in the mine. The hazard prediction module is used to accurately predict the dangerous area of the mine through the analysis of the deformation area. The mine mountain surface data collected by humans and machines can more accurately predict the areas where surface depression is dangerous. That is, without hardware in contact with the ground, more accurate early warning of areas where surface depression may be dangerous can be achieved. To ensure the safe production of mines.

Embodiment 2: Based on Embodiment 1, a mine safety risk early warning and prediction system and image modeling module include:

The image acquisition submodule is used to collect images of the area where the mine is located based on the shooting equipment set up on the drone, and obtain multiple images of the mine mountain;

The image conversion sub-module is used to input multiple mine mountain images into the preset plane conversion model to obtain corresponding splicable images;

The graphics fusion sub-module is used to frequency decompose all splicable images and then fuse them to obtain a fused image, and build a model based on the fused image to obtain a preliminary mountain structure model.

In this embodiment, the preset plane transformation model utilizes a large number of image transformation instances in advance (that is, it includes images in other coordinates that are different from the world coordinate system and corresponding instances of converting them into images in the world coordinate system). The trained model can convert images in different coordinate systems that are different from the world coordinate system into images that can be spliced (such as images in the world coordinate system).

In this embodiment, the splicable images are images in the same coordinate system that are output after passing through a preset transformation model.

The beneficial effects of the above technology are: based on the preset plane conversion model, multiple mine mountain images in different coordinate systems can be converted into splicable images in the same coordinate system, thereby realizing coordinate conversion of the mine images and facilitating subsequent image fusion steps. , and further, the stitchable images obtained after all coordinate transformations are fused to obtain a preliminary mountain structure model.

Embodiment 3: Based on Embodiment 1, a mine safety risk early warning and prediction system, graphics fusion sub-module, includes:

The image conversion unit is used to perform Fourier transform on the splicable image, obtain the spectrogram, and determine the brightness value of each pixel in the spectrogram as the frequency component representation value in the splicable image;

The area determination unit is used to determine the same frequency component representation value in different splicable images, and the area surrounded by pixels corresponding to all the same frequency component representation values in each two splicable images is regarded as corresponding to two Stitchable areas between stitchable images;

The image splicing unit is used to splice all splicable images in pairs based on all splicable areas to obtain a fused image;

The model building unit is used to extract three-dimensional point clouds from the fused image, conduct modeling based on the three-dimensional point cloud data, and obtain a preliminary mountain structure model.

In this embodiment, the frequency component representation value is a numerical value that represents the size of the frequency component at the corresponding pixel point.

In this embodiment, the splicable area is the partial image of the same part of the mine mountain structure included in the two splicable images.

In this embodiment, the fused image is an image obtained by merging all splicable images in pairs.

In this embodiment, the technology for extracting three-dimensional point clouds from fused images can, for example, use the GraphX convolution network to convert from 2D images to three-dimensional point clouds.

In this embodiment, the three-dimensional point cloud data is the three-dimensional coordinates of the pixel at the actual mountain position, including the specific position and corresponding depth.

The beneficial effects of the above technology are: based on the same frequency components contained in different splicable images, the splicable areas containing the same part of the mine mountain structure in different splicable images are determined, and all splicable images are processed based on the splicable areas. Two splicing methods achieve deduplication and effective splicing of splicable images, and extract non-redundant three-dimensional point cloud data based on the fused image obtained by final splicing, and finally achieve the accurate construction of a preliminary mountain structure model.

Embodiment 4: Based on Embodiment 3, a mine safety risk early warning and prediction system and an image splicing unit include:

An image generation subunit, used to generate the first sub-image corresponding to each frequency component of the splicable image based on the spectrogram;

The image distinction subunit is used to determine all the first partial sub-images of the low-frequency component of the splicable area in the two splicable images to be fused (all the first partial sub-images here include two splicable areas (that is, two The splicable images each contain a splicable image) a partial image in all sub-images of the low-frequency component) and all the second-part sub-images of the high-frequency component (here all the second-part sub-images include two splicable regions in part of all sub-images of high-frequency components);

The low-frequency processing subunit is used to determine the average gray value of all pixels at the same position in all the first sub-images, and assign the average gray value to the pixels at the corresponding positions in the blank image to obtain a low-frequency fusion image. ;

The high-frequency processing subunit is used to determine the maximum gray value of all pixels at the same position in all the second part sub-images, and assign the maximum gray value to the pixels at the corresponding positions in the blank image, Obtain high-frequency fusion images;

The image splicing subunit overlays and synthesizes the low-frequency fusion image and the high-frequency fusion image to obtain a new splicable area. Based on the new splicable area, the two splicable images to be fused are spliced to obtain a fused image.

In this embodiment, the first sub-image corresponding to each frequency component of the splicable image generated based on the spectrogram is:

Determine the brightness value of each pixel in the spectrogram, and use the median of the brightness values of all pixels in the spectrogram as the division threshold, and use pixels that are not less than the division threshold as the high-frequency component of the spectrogram. And the remaining points in the spectrogram except the high-frequency components are regarded as low-frequency components;

The ratio of the ordinal number of each frequency component among all frequency components to the total number of all frequency components is regarded as the pixel decomposition ratio;

The pixel value of the pixel corresponding to the current frequency component is retained, and the product of the pixel value and the pixel decomposition ratio of the remaining pixels in the splicable image except the pixel corresponding to the current frequency component is regarded as the corresponding pixel at the current frequency. The decomposed pixel value under the component is based on the pixel value of the current frequency component in the splicable image and the decomposed pixel value of the remaining pixels in the splicable image except the pixel corresponding to the current frequency component under the current frequency component to generate the current frequency The first sub-image of the component.

In this embodiment, two splicable images to be fused are two splicable images that include splicable regions with the same frequency components.

In this embodiment, the low-frequency fusion image is a fusion image obtained by averaging the pixels of all the first sub-images of the low-frequency component in the splicable area in the two splicable images to be fused.

In this embodiment, the high-frequency fusion image is a fusion image obtained by taking the maximum pixel value of all the first sub-images of the low-frequency component in the splicable area of the two splicable images to be fused.

The beneficial effects of the above technology are: distinguishing the sub-images of all frequency components of the fused image according to high frequency and low frequency, and processing them according to different processing methods to obtain high-frequency and low-frequency fused images, which can be more accurate and easier for subsequent processing. fused images.

Embodiment 5: Based on Embodiment 1, a mine safety risk early warning and prediction system, the scope determination module includes:

The boundary optimization submodule is used to divide the fusion image that constructs the preliminary mountain structure model into multiple second sub-images of unit area, perform image boundary detection on all second sub-images, and determine the boundary parts of all second sub-images. Perform boundary enhancement on the boundary parts of all second sub-images to obtain a set of boundary-enhanced sub-images;

The correction value calculation submodule is used to determine the correction value of each boundary enhancement sub-image in the boundary enhancement sub-image set based on the following formula:

;

In the formula, is the correction value of the boundary enhancer image, a is the ordinate value of the lower boundary of the boundary enhancer image in the plane rectangular coordinate system, b is the ordinate value of the upper boundary of the boundary enhancer image in the plane rectangular coordinate system, c is the abscissa value of the left boundary of the boundary enhancer image in the plane rectangular coordinate system, d is the abscissa value of the right boundary of the boundary enhancer image in the plane rectangular coordinate system,/> For the coordinates in the boundary enhancer image/> The maximum value of the pixel difference between the pixel at and each pixel in the enhanced boundary in the corresponding boundary enhanced sub-image,/> For the coordinates in the boundary enhancer image/> The minimum value of the pixel difference between the pixel at and each pixel in the enhanced boundary in the corresponding boundary enhancement sub-image,/> Differentiate for the independent variable;

The range judgment submodule is used to regard the boundary enhancement sub-images whose correction values are greater than the rated correction value in the boundary enhancement sub-image set as areas to be corrected, and to regard the corresponding area ranges of all areas to be corrected in the preliminary mountain structure model as area corrections. scope.

In this embodiment, the second sub-image of the unit area is to divide the fused image into multiple sub-images with a length and a width of 1 decimeter.

In this embodiment, image boundary detection is based on the existing edge detection algorithm to perform boundary detection on all sub-images and determine the boundary parts of all sub-images.

In this embodiment, boundary enhancement is based on the existing neighborhood enhancement algorithm to smooth and sharpen the outline of the mine mountain in all sub-images to make it more obvious.

In this embodiment, the boundary-enhanced sub-image set is a combination of all boundary-enhanced sub-images.

In this embodiment, the correction value may represent the true representation of the boundary enhancer image on its corresponding mountain surface.

In this embodiment, the rated correction value is the critical value of the boundary enhancer image's true representation of the corresponding mountain surface. When the correction value is greater than the rated correction value, it means that the boundary enhancer image cannot truly reflect the corresponding mountain surface. Set as the area to be corrected, otherwise the boundary enhancer image can truly reflect the corresponding mountain surface.

The beneficial effects of the above technology are: segment the fused image, and analyze the segmented sub-images to determine whether they are areas to be corrected, which reduces the process and speeds up the analysis and processing.

Embodiment 6: Based on Embodiment 1, a mine safety risk early warning and prediction system, laser modeling module, includes:

The laser sampling sub-module determines multiple scanning points in the regional correction range, controls the drone to scan at the actual position corresponding to the scanning point in the mine mountain, and obtains three-dimensional point cloud data of the corresponding scanning point;

The coincidence degree calculation sub-module is used to calculate the coincidence degree of the three-dimensional point cloud data of all adjacent scanning points;

The supplementary scanning sub-module is used to insert new scanning points between adjacent scanning points whose coincidence degree is less than the minimum coincidence degree threshold, and to supplementally obtain the three-dimensional point cloud data of the new scanning points until all adjacent scans are newly determined. When the degree of coincidence between the three-dimensional point cloud data of the points is not less than the minimum coincidence degree threshold, the first three-dimensional point cloud data is obtained;

The point cloud deduplication submodule is used to randomly remove a group of three-dimensional point cloud data of adjacent scanning points whose coincidence degree exceeds the maximum coincidence degree threshold in the first three-dimensional point cloud data, and obtain the second three-dimensional point cloud data;

A registration point sub-module, used to determine the same three-dimensional point cloud in the three-dimensional point cloud data of adjacent scanning points in the second three-dimensional point cloud data as the registration point corresponding to the adjacent scanning point;

The point cloud synthesis submodule is used to synthesize the three-dimensional point cloud data corresponding to adjacent scanning points based on the registration points to obtain a complete point cloud of the regional correction range;

The model construction submodule is used to build a regional correction range model based on the complete point cloud of the regional correction range.

In this embodiment, the scanning point is set as the starting point based on any point on the edge of the area correction range, and every 10 meters is set as the scanning point until the final interval is less than 10 meters.

In this embodiment, the degree of coincidence is the percentage of the total number of identical three-dimensional point cloud data in all three-dimensional point cloud data of adjacent scanning points to the sum of all three-dimensional point cloud data of two scanning points.

In this embodiment, the minimum coincidence threshold is the minimum coincidence degree of the three-dimensional point cloud data of adjacent scanning points. If it is less than the minimum coincidence threshold, it means that the distance between the corresponding two scanning points is too large and the data coincidence degree is too low. It needs to be in the middle. Insert new scan points.

In this embodiment, the first three-dimensional point cloud data is to insert a new scan point and obtain the corresponding three-dimensional point cloud data to ensure that the coincidence degree between the three-dimensional point cloud data of all adjacent scan points is not less than the minimum coincidence threshold. 3D point cloud data.

In this embodiment, the maximum coincidence threshold is the maximum coincidence degree of the three-dimensional point cloud data of adjacent scanning points. If it is greater than the maximum coincidence threshold, it means that the distance between the corresponding two scanning points is too small and the data overlap is too high. It is necessary to compare the two scanning points. Randomly remove one group from the group of scan point data.

In this embodiment, the second three-dimensional point cloud data is all three-dimensional point cloud data except for a random set of three-dimensional point cloud data in adjacent scan points whose coincidence degree exceeds the maximum coincidence threshold in the first three-dimensional point cloud data.

In this embodiment, the registration points are the same three-dimensional point cloud in the three-dimensional point cloud data of adjacent scanning points in the second three-dimensional point cloud data, and the point cloud synthesis is performed relying on the registration points as reference points.

The beneficial effects of the above technology are: scanning multiple scanning points with a drone to obtain more accurate three-dimensional point cloud data, and determining whether new scanning points need to be inserted or determined by judging the degree of overlap of the three-dimensional point cloud data of adjacent scanning points. It is the three-dimensional point cloud data that removes the scanned points, which is faster and saves response time. The point cloud is synthesized based on the registration points, and the complete point cloud of the regional correction range is obtained more conveniently and quickly.

Embodiment 7: Based on Embodiment 1, a mine safety risk early warning and prediction system and an orientation determination module include:

The size unification submodule is used to unify the size of the preliminary mountain structure model and regional correction range model;

The position determination submodule is used to determine the fusion orientation of the regional correction range model in the preliminary mountain structure model.

In this embodiment, the size is unified and the preliminary mountain model and the regional correction range model are processed into a size of 750 pixels * 422 pixels, confirming that they can be directly spliced and merged.

In this embodiment, the fusion orientation is the position where the regional correction model is fused in the preliminary mountain structure model, and model splicing is performed based on the fused position.

The beneficial effect of the above technology is that the preliminary mountain model and the regional correction range model can be processed and merged at the same scale, making the fusion of the models more convenient and simple.

Embodiment 8: Based on Embodiment 1, a mine safety risk early warning and prediction system and an area determination module include:

The fusion sub-module fuses the regional correction range model and the preliminary mountain structure model based on the fusion orientation to obtain a complete mountain structure model;

The deformation determination submodule is used to compare the complete mountain structure model with the preset mountain structure model to determine the deformation area.

In this embodiment, fusing the regional correction range model and the preliminary mountain structure model based on the fusion orientation to obtain a complete mountain structure model includes: removing the part of the area to be corrected in the preliminary mountain structure model, and directly splicing the regional correction model to the preliminary mountain structure model. , to obtain a complete mountain structure model.

In this embodiment, the complete mountain structure model is compared with the preset mountain structure model to determine the deformation area, including:

The different areas in the complete mountain structure model and the preset mountain structure model are regarded as deformation areas.

The beneficial effect of the above technology is that the deformation area can be more conveniently and accurately determined by comparing the complete mountain structure model with the preset mountain structure model.

Embodiment 9: Based on Embodiment 1, a mine safety risk early warning and prediction system and a risk prediction module include:

The initial curvature calculation submodule is used to calculate the depression curvature of the deformation area based on the preset mountain structure model to obtain the initial curvature of the deformation area;

The depression curvature submodule is used to calculate the depression curvature based on the structural changes of the complete mountain structure model in the deformation area to obtain the current curvature of the deformation area;

The prediction submodule is used to predict the deformation area as a dangerous area of the mine when the difference between the current curvature and the initial curvature is greater than the preset value.

In this embodiment, the preset mountain structure model is a mountain structure model built by using equipment to collect data in advance (when no structural deformation occurs in the mine mountain).

In this embodiment, the concave curvature is calculated by analyzing the preset mountain structure model part corresponding to the position of the deformation region based on the second basic formula of the existing curved surface, so that the position of any point in the deformation region (set as point a) can be The curvature of is equal to the curvature of the tangent vector direction curve at its position, and then the mathematical relationship between the curvature of point a in the deformation area and Gaussian curvature is determined based on Euler's formula (the curvature of any tangent direction passing through surface point a satisfies Euler's formula). According to Gaussian The mathematical relationship between curvature (Gaussian curvature is a real symmetric matrix, then it can be decomposed into the product of an orthogonal matrix and a diagonal matrix) and the curvature of point a in the deformation area can be obtained by performing the above process on all points in the deformation area. Traverse the curvature of all points in the deformation area, and set the maximum curvature as the concave curvature of the deformation area.

In this embodiment, the initial curvature is the concave curvature of the preset mountain structure model part corresponding to the position of the deformation region, which is set as the initial curvature.

In this embodiment, the current curvature is the concave curvature of the complete mountain structure model part corresponding to the position of the deformation region, which is set as the current curvature.

In this embodiment, the preset value is a value calculated and summarized in advance through a large amount of data. When the difference between the current curvature and the initial curvature is greater than the preset value, it means that the depression in the deformation area of the mine corresponding to the current curvature is irreversible, and the mine The curvature of the deformed region will gradually increase, and the deformed region will collapse.

The beneficial effects of the above technology are: relying on the curvature calculation results, it can predict whether the deformation area is a dangerous area of the mine, improve the prediction efficiency, realize the advance warning of mine depressions, and leave more time for the maintenance or early warning prevention of mine depressions. .

Embodiment 10: Based on Embodiment 1, a mine safety risk early warning and prediction system and hazard early warning module include:

The signal generation submodule is used to generate corresponding early warning signals based on the location of the mine's dangerous areas;

The signal sending submodule is used to send early warning signals to the monitoring and early warning background.

In this embodiment, the early warning signal is an alarm signal containing location information.

In this embodiment, the monitoring and early warning background is the mine's monitoring room background.

The beneficial effects of the above technology are: generating corresponding early warning signals based on the location of dangerous areas in mines, providing timely early warning of dangers, and achieving advance warning of safety risk accidents caused by possible depressions and deformations in mines.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the invention. In this way, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims (8)

1. A mine safety risk early warning and prediction system, which is characterized by including: The image modeling module is used to frequency decompose and fuse the plane-converted mine mountain images to obtain the fused image, and build a preliminary mountain structure model based on the fused image; The range determination module is used to perform unit area decomposition and boundary enhancement on the fused image, obtain the boundary enhancement sub-image, and determine the regional correction range based on the correction value of the boundary enhancement sub-image; The laser modeling module is used to re-model the regional correction range and obtain the regional correction range model; The orientation determination module is used to process based on the preliminary mountain model and the regional correction range model, and determine the fusion orientation of the regional correction range model in the preliminary mountain structure model; The area determination module is used to obtain a complete mountain structure model based on the fusion orientation, and determine the deformation area of the mine through the complete mountain structure model and the preset mountain structure model; The hazard prediction module is used to calculate the depression curvature of the deformation area of the mine, obtain the current curvature of the deformation area, and predict the dangerous area of the mine based on the current curvature and the corresponding initial curvature of the deformation area in the preset mountain structure model; Danger early warning module, used to send early warning signals based on the location of dangerous areas in mines; Scoping modules include: The boundary optimization submodule is used to divide the fusion image that constructs the preliminary mountain structure model into multiple second sub-images of unit area, perform image boundary detection on all second sub-images, and determine the boundary parts of all second sub-images. Perform boundary enhancement on the boundary parts of all second sub-images to obtain a set of boundary-enhanced sub-images; The correction value calculation submodule is used to determine the correction value of each boundary enhancement sub-image in the boundary enhancement sub-image set based on the following formula: In the formula, θ is the correction value of the boundary enhancer image, a is the ordinate value of the lower boundary of the boundary enhancer image in the plane rectangular coordinate system, and b is the ordinate value of the upper boundary of the boundary enhancer image in the plane rectangular coordinate system. Coordinate value, c is the abscissa value of the left boundary of the boundary enhancer image in the plane rectangular coordinate system, d is the abscissa value of the right boundary of the boundary enhancer image in the plane rectangular coordinate system, P (x, y) is The maximum value of the pixel difference between the pixel at coordinates (x, y) in the boundary enhancement sub-image and each pixel in the enhancement boundary in the corresponding boundary enhancement sub-image, Q(x, y) is the boundary enhancer The minimum value of the pixel difference between the pixel at coordinates (x, y) in the image and each pixel in the enhanced boundary in the corresponding boundary enhancement sub-image, dx and dy are independent variable differentials; The range judgment submodule is used to regard the boundary enhancement sub-images whose correction values are greater than the rated correction value in the boundary enhancement sub-image set as areas to be corrected, and to regard the corresponding area ranges of all areas to be corrected in the preliminary mountain structure model as area corrections. scope; Laser modeling module, including: The laser sampling sub-module determines multiple scanning points in the regional correction range, controls the drone to scan at the actual position corresponding to the scanning point in the mine mountain, and obtains three-dimensional point cloud data of the corresponding scanning point; The coincidence degree calculation sub-module is used to calculate the coincidence degree of the three-dimensional point cloud data of all adjacent scanning points; The supplementary scanning sub-module is used to insert new scanning points between adjacent scanning points whose coincidence degree is less than the minimum coincidence degree threshold, and to supplementally obtain the three-dimensional point cloud data of the new scanning points until all adjacent scans are newly determined. When the degree of coincidence between the three-dimensional point cloud data of the points is not less than the minimum coincidence degree threshold, the first three-dimensional point cloud data is obtained; The point cloud deduplication submodule is used to randomly remove a group of three-dimensional point cloud data of adjacent scanning points whose coincidence degree exceeds the maximum coincidence degree threshold in the first three-dimensional point cloud data, and obtain the second three-dimensional point cloud data; A registration point sub-module, used to determine the same three-dimensional point cloud in the three-dimensional point cloud data of adjacent scanning points in the second three-dimensional point cloud data as the registration point corresponding to the adjacent scanning point; The point cloud synthesis submodule is used to synthesize the three-dimensional point cloud data corresponding to adjacent scanning points based on the registration points to obtain a complete point cloud of the regional correction range; The model construction submodule is used to build a regional correction range model based on the complete point cloud of the regional correction range. 2. A mine safety risk early warning and prediction system according to claim 1, characterized in that the image modeling module includes: The image acquisition submodule is used to collect images of the area where the mine is located based on the shooting equipment set up on the drone, and obtain multiple images of the mine mountain; The image conversion sub-module is used to input multiple mine mountain images into the preset plane conversion model to obtain corresponding splicable images; The graphics fusion sub-module is used to frequency decompose all splicable images and then fuse them to obtain a fused image, and build a model based on the fused image to obtain a preliminary mountain structure model. 3. A mine safety risk early warning and prediction system according to claim 2, characterized in that the graphics fusion sub-module includes: The image conversion unit is used to perform Fourier transform on the splicable image, obtain the spectrogram, and determine the brightness value of each pixel in the spectrogram as the frequency component representation value in the splicable image; The area determination unit is used to determine the same frequency component representation value in different splicable images, and the area surrounded by pixels corresponding to all the same frequency component representation values in each two splicable images is regarded as corresponding to two Stitchable areas between stitchable images; The image splicing unit is used to splice all splicable images in pairs based on all splicable areas to obtain a fused image; The model building unit is used to extract three-dimensional point clouds from the fused image, conduct modeling based on the three-dimensional point cloud data, and obtain a preliminary mountain structure model. 4. A mine safety risk early warning and prediction system according to claim 3, characterized in that the image splicing unit includes: An image generation subunit, used to generate the first sub-image corresponding to each frequency component of the splicable image based on the spectrogram; The image distinction subunit is used to determine all the first sub-images of the low-frequency component and all the second sub-images of the high-frequency component in the splicable area in the two splicable images to be fused; The low-frequency processing subunit is used to determine the average gray value of all pixels at the same position in all the first sub-images, and assign the average gray value to the pixels at the corresponding positions in the blank image to obtain a low-frequency fusion image. ; The high-frequency processing subunit is used to determine the maximum gray value of all pixels at the same position in all the second part sub-images, and assign the maximum gray value to the pixels at the corresponding positions in the blank image, Obtain high-frequency fusion images; The image splicing subunit overlays and synthesizes the low-frequency fusion image and the high-frequency fusion image to obtain a new splicable area. Based on the new splicable area, the two splicable images to be fused are spliced to obtain a fused image. 5. A mine safety risk early warning and prediction system according to claim 1, characterized in that the orientation determination module includes: The size unification submodule is used to unify the size of the preliminary mountain structure model and regional correction range model; The position determination submodule is used to determine the fusion orientation of the regional correction range model in the preliminary mountain structure model. 6. A mine safety risk early warning and prediction system according to claim 1, characterized in that the area determination module includes: The fusion sub-module fuses the regional correction range model and the preliminary mountain structure model based on the fusion orientation to obtain a complete mountain structure model; The deformation determination submodule is used to compare the complete mountain structure model with the preset mountain structure model to determine the deformation area. 7. A mine safety risk early warning and prediction system according to claim 1, characterized in that the risk prediction module includes: The initial curvature calculation submodule is used to calculate the initial curvature of the deformation area in the corresponding range of the preset mountain structure model; The depression curvature submodule is used to calculate the current curvature of the deformation area in the complete mountain structure model; The prediction sub-module is used to predict the deformation area where the difference between the corresponding current curvature and the initial curvature is greater than the preset value as a mine dangerous area. 8. A mine safety risk early warning and prediction system according to claim 1, characterized in that the danger early warning module includes: The signal generation submodule is used to generate corresponding early warning signals based on the location of the mine's dangerous areas; The signal sending submodule is used to send early warning signals to the monitoring and early warning background.