CN-122023487-A - Coal surface reflectivity enhancement method based on multi-angle spectrum consistency constraint

CN122023487ACN 122023487 ACN122023487 ACN 122023487ACN-122023487-A

Abstract

The invention relates to the technical field of spectral reflectivity, in particular to a coal surface reflectivity enhancement method based on multi-angle spectral consistency constraint, which comprises the steps of synchronously acquiring multi-angle spectral data by a plurality of unmanned aerial vehicles, and constructing a spatial registration image group, extracting consistent pixels to construct an angle reflection structure, identifying reflection difference wave bands and directions, executing directivity correction, and fusing to generate an enhanced reflectivity image result. According to the invention, a multi-angle spectral reflectance image group is acquired through a plurality of unmanned aerial vehicles, a pixel corresponding relation is established by combining a time mark and space registration, reflectance values at the same position are polymerized according to angles to form a directional structure, an abnormal wave band area is identified through difference value calculation, angle differences are processed by combining spectral sequence rearrangement and direction correction, a unified reflectance value set is constructed, a two-dimensional image matrix is generated through space splicing and fusion, and the spatial consistency and the direction adaptability of multi-angle spectral data of a coal area are enhanced.

Inventors

KANG WENWEN
YANG GONGBING
REN HAITAO
HE HANG
LIU XIAOGUANG
WANG XIULIANG
LI BIN
ZHANG WEI

Assignees

国能(天津)港务有限责任公司

Dates

Publication Date: 20260512
Application Date: 20260202

Claims (10)

1. The coal surface reflectivity enhancement method based on multi-angle spectrum consistency constraint is characterized by comprising the following steps of: S1, scanning the surface of a coal pile along different heading by controlling a plurality of unmanned aerial vehicles, collecting spectral reflection data in a coverage area, recording shooting azimuth and flight coordinates, registering images and performing track matching processing, and constructing a coal area reflectivity image group under a multi-angle corresponding relation; s2, extracting coal surface area points with position consistency based on the coal area reflectivity image group under the multi-angle corresponding relation, grouping image pixel points at the same spatial position according to the observation angle, establishing a pairing relation of the direction and the reflectivity value, and constructing a coal surface pixel and angle reflection data structure; S3, based on the aggregation structure of the coal surface pixels and the angle reflection data, carrying out sequence expansion and difference calculation on the reflectivity values under different angles, identifying a band region with large variation amplitude, marking a corresponding observation direction, and constructing a coal surface reflectivity directivity difference region identification result; And S4, extracting a related reflectivity sequence from the original image according to the abnormal observation direction in the identification result of the coal surface reflectivity directivity difference region, and carrying out offset correction and trend sequencing on the difference wave bands to generate an angle correction reflectivity numerical value set.
2. The method for enhancing the reflectivity of the coal surface based on the multi-angle spectrum consistency constraint of claim 1, wherein the coal region reflectivity image group comprises multi-view image data, a reflectivity distribution matrix and space-time identification information, an aggregation structure of image elements of the coal surface and angle reflection data comprises angle number information, space corresponding image elements and a direction reflectivity set, a coal surface reflectivity directivity difference region identification result comprises a high difference wave band interval, an abnormal observation direction mark and directivity change characteristics, and an angle corrected coal region reflectivity numerical value set comprises an adjusted spectrum sequence, a difference wave band correction value and unified direction reflectivity data.
3. The method for enhancing the reflectivity of the coal surface based on the multi-angle spectrum consistency constraint of claim 1, wherein the step of obtaining S1 is as follows: s101, extracting continuous spectrum reflection data in a coverage area frame by frame based on spectrum image frame sequences acquired in the flight process of a plurality of unmanned aerial vehicles along different airlines and attitude angles, combining flight path coordinate sequences synchronously acquired by the unmanned aerial vehicles with shooting azimuth angle data, corresponding the image frames with corresponding time sequence numbers in the processing process, and obtaining a space-time mapping data set corresponding to the spectrum image frames through corresponding processing of the image frames and position information thereof; S102, according to the image time sequence numbers and the path coordinate sequences in the space-time mapping data set corresponding to the spectral image frames, executing a pairing process according to the sequence of the image frames, correspondingly processing path coordinate changes among adjacent image frames and shooting azimuth angle data, and comparing the continuity differences of the image frames with the attitude angle data to obtain a displacement association information group of the multi-view image frames; and S103, according to the coordinate relation content in the displacement association information group of the multi-view image frames, readjusting the image frame content under the unified space parameter according to the relation between the path numerical value and the attitude angle numerical value between the adjacent image frames, so that the reflection data of each image frame are correspondingly processed under the same coordinate system, and finally, the coal region reflectivity image group under the multi-angle corresponding relation is obtained.
4. The method for enhancing the reflectivity of the coal surface based on multi-angle spectrum consistency constraint of claim 3, wherein the method comprises the following steps: The coordinate relation content in the displacement association information group of the multi-view image frames is used for screening image frame pairs meeting that the change amplitude of the path coordinates and the change amplitude of the shooting azimuth angles are both in a preset threshold range so as to establish an effective matching relation; the process of readjusting the content of the image frames under the unified space parameters comprises the steps of carrying out normalization processing on the space positions of the image frames based on the effective matching relation, and carrying out image reprojection operation according to the change of the attitude angles among the image frames so as to realize space alignment; the coal region reflectivity image group under the multi-angle corresponding relation is obtained through image frame reconstruction under a unified coordinate system and comprises a plurality of image frames with different shooting azimuth angles, wherein the image frames meet the effective matching relation.
5. The method for enhancing the reflectivity of the coal surface based on the multi-angle spectrum consistency constraint of claim 1, wherein the step of obtaining S2 is as follows: S201, based on each angle image frame in the coal region reflectivity image group under the multi-angle corresponding relation, cross screening is carried out on position coordinates of pixel points at the same position in the image frames under different angles, spatial reference coordinate values of each image frame are called, alignment comparison operation is carried out on spatial position index values of all the pixel points, and a position point spatial consistency index set is obtained; S202, according to the coordinate numbers of all the position points in the position point space consistency index set, performing classification operation on the pixel points with the same coordinate numbers according to the acquisition sequence of the spectrogram frames, calling different angle number values to perform corresponding assignment processing on the reflectivity data corresponding to all the angles, and obtaining a reflectivity data set under the angle sequence; s203, carrying out structure merging processing on each group of direction numbers and corresponding reflectivity numbers based on the corresponding space coordinate numbers of each group according to the reflectivity numbers and direction number data in the reflectivity data groups under the angle sequence, and carrying out unified mapping processing on all space position index numbers to obtain a coal surface pixel angle reflection data structure.
6. The method for enhancing the reflectivity of the coal surface based on the multi-angle spectrum consistency constraint of claim 1, wherein the step of obtaining S3 is as follows: S301, extracting numerical sequences of the reflectivity value sequences under each group of observation directions based on the reflectivity value sequences corresponding to each position index in the coal surface pixel angle reflection data structure, and calling sequence order data corresponding to each direction to perform position comparison processing to obtain a reflectivity sequence set corresponding to each wave band; S302, extracting the reflectivity values of the same wave band in each observation direction according to the wave band corresponding to the same wave band position in the reflectivity sequence set, calling the direction index sequence to carry out pairwise comparison on the values in the adjacent observation directions, and numbering the wave band position sequences after each group of comparison according to the direction sequence to obtain a direction sequence wave band comparison result set; S303, monitoring the numerical value variability degree in the direction sequence under each group of wave bands according to the numerical value variability state of each wave band under different observation directions in the direction sequence wave band comparison result set, extracting according to the direction index corresponding to the variability change range, classifying the wave bands with prominent variability range and the corresponding contents of the direction indexes, and obtaining the recognition result of the coal surface reflectivity directivity difference region.
7. The method for enhancing the reflectivity of the coal surface based on the multi-angle spectrum consistency constraint of claim 1, wherein the step of obtaining S4 is as follows: S401, extracting a spectral reflectance sequence in a corresponding direction from an original reflectance image frame according to an observation direction index contained in a recognition result of the directional difference region of the reflectance of the coal surface, calling spectral values of each band position in the difference region, and positioning the content of the corresponding image region by combining spatial index data to obtain an abnormal directional reflectance band sequence; S402, calling the sequence of the spectrum values of the same position in other direction sequences according to the spectrum value of each band position in the abnormal direction reflectivity band sequence, performing comparison processing on the numerical relationship of the spectrum values, and performing position adjustment operation on the spectrum value sequence according to the sequence offset to obtain a band position comparison adjustment result; S403, according to the sequence of the adjusted spectrum values in the band position comparison adjustment result, the band position comparison adjustment result is in one-to-one correspondence with the spatial indexes in the original sequence, the content of the reflectivity data frame is classified and arranged according to the serial numbers of the spatial positions, and the values in all abnormal observation directions are reclassified into the corresponding spatial mapping indexes, so that an angle correction reflectivity value set is obtained.
8. The multi-angle spectral consistency constraint-based coal surface reflectivity enhancement method of claim 7, wherein: when the spectrum value of each band position in the abnormal direction reflectivity band sequence is called to carry out comparison processing in the sequence of spectrum values of the same position in other direction sequences, position adjustment operation is only carried out on the spectrum values meeting the sequence difference which does not exceed the preset adjustment tolerance by setting the limiting condition of the sequence offset range; In the process that the sequence of the adjusted spectrum values in the band position comparison adjustment result corresponds to the spatial index in the original sequence one by one, the image area content with continuous spatial numbers is preferentially selected for sorting on the premise of matching the continuity of the spatial index; And when each spectrum value in the angle correction reflectivity value set is classified into a corresponding space mapping index, performing normalization processing according to spectrum value distribution trends of sequences in different directions under the same space position so as to eliminate interference of the difference of the observation directions on reflectivity data.
9. The method for enhancing the reflectivity of the surface of coal based on multi-angle spectral consistency constraint of claim 1, further comprising: s5, based on the angle-corrected coal region reflectivity value set, splicing the reflectivity value of each position into an image matrix, fusing all direction data and mapping the image matrix to space coordinates, and constructing a coal surface reflectivity enhancement image result; The coal surface reflectivity enhancement image result comprises a fused reflectivity image, a corrected image matrix and a space coordinate mapping result.
10. The method for enhancing the reflectivity of the coal surface based on the multi-angle spectrum consistency constraint of claim 9, wherein the step of S5 is: S501, extracting spectral value content of each position point according to coordinate sequence based on the reflectivity values corresponding to each spatial index in the angle correction reflectivity value set, and calling a spatial coordinate sequence to splice the reflectivity of each position point, and arranging all the reflectivity values according to a two-dimensional structure to obtain a spatial spliced image matrix; S502, synchronously processing image data of the same area in different observation directions according to the image pixel content of each spatial position in the spatial stitching image matrix, sequentially extracting the data frame content of the corresponding position in the image matrix in each direction, and calling the reflectivity value in each direction to perform channel-level corresponding processing to obtain a multi-directional reflectivity stitching matrix; s503, extracting pixel positions corresponding to each spatial index for the image data frame in the multi-directional reflectivity splicing matrix, comparing each pixel position in the image to the index number of the surface area of the coal pile through a spatial parameter mapping rule, filling the image data content according to the corresponding relation, and obtaining a coal surface reflectivity enhancement image result.

Description

Coal surface reflectivity enhancement method based on multi-angle spectrum consistency constraint Technical Field The invention relates to the technical field of spectral reflectivity, in particular to a coal surface reflectivity enhancement method based on multi-angle spectral consistency constraint. Background The technical field of spectral reflectivity comprises an imaging detection method for acquiring, modeling and analyzing reflection characteristics of a target surface under different illumination and observation angles, is widely applied to the fields of geological investigation, environmental remote sensing, agricultural monitoring, energy resource evaluation and the like, and has the core content of realizing accurate estimation and analysis of the surface reflectivity of the earth surface or an object by combining technical means such as radiation calibration, geometric correction, reflectivity inversion and consistency modeling based on hyperspectral or multispectral imaging equipment. In a coal detection scene, the spectral reflectivity measurement needs to cope with the influence of factors such as external environment illumination interference, observation angle change, target surface roughness and the like, so that the stability and applicability of reflectivity data are improved by combining means such as multi-angle imaging, reflectivity interpolation modeling, angle normalization reconstruction and the like in the field. The method for enhancing the reflectivity of the coal surface based on the multi-angle spectrum consistency constraint is characterized in that in the detection of the spectrum of the coal surface, the reflectivity enhancement processing is carried out by introducing a spectrum consistency constraint model aiming at the problem of distortion of reflectivity data under the multi-angle observation condition. According to the method, the spectrum data obtained by multi-angle repeated observation is utilized to construct an angle-dependent differential expression model, the spectrum difference between observation angles is fitted and corrected in a numerical optimization mode by combining the change trend of the reflectivity, and the multi-frame reflectivity mapping is completed by adopting a reference view reflectivity reconstruction mechanism. The core means comprises angle differential modeling, multi-frame spectrum sequence joint analysis, spectrum angle dependence weight fitting and consistency reconstruction calculation, and is used for improving the stability of the reflectivity data of the coal surface in an outdoor complex environment. In the prior art, the reflectivity is modeled by single-point observation and image sequence processing, continuous correlation of multi-angle observation data in a large range in space-time is difficult to obtain, data matching is inaccurate due to lack of time marks and spatial registration support among images, limitation exists in angle reflection characteristic expression, deviation of reflectivity data of the same spatial position under different angles is easy to occur, positioning and correction capability of a local distortion area are not provided, and the requirement of stable reflectivity extraction under the condition of strong angle interference or severe reflection change is difficult to meet. Disclosure of Invention In order to solve the technical problems in the prior art, the embodiment of the invention provides a coal surface reflectivity enhancement method based on multi-angle spectrum consistency constraint. The technical scheme is as follows: The coal surface reflectivity enhancement method based on multi-angle spectrum consistency constraint comprises the following steps: S1, scanning the surface of a coal pile along different heading by controlling a plurality of unmanned aerial vehicles, collecting spectral reflection data in a coverage area, recording shooting azimuth and flight coordinates, registering images and performing track matching processing, and constructing a coal area reflectivity image group under a multi-angle corresponding relation; s2, extracting coal surface area points with position consistency based on the coal area reflectivity image group under the multi-angle corresponding relation, grouping image pixel points at the same spatial position according to the observation angle, establishing a pairing relation of the direction and the reflectivity value, and constructing a coal surface pixel and angle reflection data structure; S3, based on the aggregation structure of the coal surface pixels and the angle reflection data, carrying out sequence expansion and difference calculation on the reflectivity values under different angles, identifying a band region with large variation amplitude, marking a corresponding observation direction, and constructing a coal surface reflectivity directivity difference region identification result; And S4, extracting a related reflectivity sequence from the ori