Parametric 3d point cloud simulation generation method for round wall measuring instrument

Abstract

A kind of parameterized three-dimensional point cloud simulation generation method for circular wall measuring instrument, and the method steps are as follows: (1) read the hexadecimal encoding data stream output by circular wall measuring instrument, check and parse to obtain integer code value sequence;(2) restore integer code value sequence to voltage sequence, denoising and missing section repair to obtain normalized sequence;(3) extract periodic event point sequence, and construct phase parameter in each circle interval;(4) phase uniform resampling is carried out to each circle signal, to construct two-dimensional development matrix;(5) two-dimensional development matrix is parameterized three-dimensional mapping according to circumferential angle parameter, axial position parameter and radial mapping function, generates point coordinate and intensity attribute;(6) output three-dimensional point cloud simulation data.The present application can convert one-dimensional time series voltage data obtained in the process of circular wall measurement into three-dimensional point cloud simulation data with structure preservation and parameterized description, to support subsequent three-dimensional structure analysis, algorithm verification and visual display.

Description

Technical Field

[0001] This invention relates to the field of circular wall measuring instruments, and in particular to a parametric three-dimensional point cloud simulation generation method for circular wall measuring instruments. Background Technology

[0002] In scenarios such as industrial internal wall inspection, cylindrical component inspection, and equipment health monitoring, circular wall measuring instruments are used to continuously sample the surface of the measured circular wall along the circumferential or axial direction. Due to limitations in cost, size, and power consumption, related measurement systems typically use one-dimensional physical optical sensors to acquire voltage signals that change over time. To adapt to embedded communication protocols and reduce transmission bandwidth and storage overhead, these voltage signals are transmitted or stored in hexadecimal encoding format.

[0003] Since circular wall measuring instruments lack 3D imaging hardware such as lidar, structured light, or multi-view vision, their output is typically only a one-dimensional time-series signal, making it difficult to directly obtain true 3D point cloud data characterizing the spatial structure of a circular wall. Current engineering practices primarily process such one-dimensional signals using waveform, spectrum, or 2D visualization methods, lacking a unified method and data generation mechanism specifically designed for circular wall detection tasks that can convert one-dimensional hexadecimal voltage data into structure-preserving parametric 3D point cloud simulation data.

[0004] Furthermore, during actual data acquisition, issues such as rotational speed drift, uneven propulsion, and trigger jitter often occur, resulting in inconsistent sampling points in each revolution. If a fixed width W is used to rearrange and unfold the data according to the sampling order, circumferential phase slippage and misalignment are likely to occur, leading to artifacts such as diagonal lines, breaks, and trailing shadows in the two-dimensional unfolded matrix and three-dimensional point cloud. At the same time, traditional pseudo-point clouds usually only output coordinates and intensity, lacking explicit identification of "repair points" such as anomaly codes, interpolation repair, and boundary fusion, which reduces the engineering usability of point clouds in algorithm evaluation and system integration. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to address the shortcomings of the prior art by providing a parametric 3D point cloud simulation generation method for circular wall measuring instruments. This method can convert one-dimensional time-series voltage data obtained during circular wall measurement into structure-preserving and parametrically described 3D point cloud simulation data without adding expensive 3D imaging hardware such as lidar, structured light, or multi-view vision, so as to support subsequent 3D structural analysis, algorithm verification, and visualization.

[0006] The technical problem to be solved by this invention is achieved through the following technical solution. This invention is a parametric 3D point cloud simulation generation method for circular wall measuring instruments, and the steps of this method are as follows:

[0007] (1) Read the hexadecimal encoded data stream raw_hex output by the circular wall measuring instrument, verify the data frame and parse it to obtain the integer code value sequence. And mark the exception codes;

[0008] (2) The obtained integer code value sequence Restored to voltage sequence :

[0009]

[0010] in For proportionality coefficient, Zero bias;

[0011] Then, denoising and missing segment repair were performed, and the voltage sequence was normalized to [0,1] to obtain u[k]:

[0012]

[0013] in >0 represents a very small constant, thus avoiding division by zero operations;

[0014] (3) In Extracting periodic event point sequences This is used to define the signal interval for each revolution and to construct phase parameters within each interval. ;

[0015] (4) Perform uniform phase resampling on each loop of signal using a uniform number of phase samples W to construct a two-dimensional expansion matrix. Periodic boundary treatment is applied to the column direction to ensure circumferential closure consistency, and uniform phase sampling points are constructed for each revolution.

[0016]

[0017]

[0018] in For linear interpolation or spline interpolation operators;

[0019] (5) Expand the two-dimensional matrix According to circumferential angle parameters Axial position parameters With radial mapping function Perform parametric 3D mapping to generate point coordinates (x, y, z) and intensity attribute I;

[0020] (6) In the process of frame verification, abnormal code marking, missing segment repair, filtering and circumferential closure processing, a quality factor is formed and fused to generate a point-level uncertainty attribute U, and finally the output is three-dimensional point cloud simulation data containing (x,y,z,I,U).

[0021] The technical problem to be solved by the present invention can also be further achieved through the following technical solution: For the parametric three-dimensional point cloud simulation generation method for circular wall measuring instrument described above, the periodic event points are obtained by local peak detection, zero-crossing detection or detection related to template signal, and minimum interval constraints and amplitude thresholds are set during the detection process to suppress glitch events.

[0022] The technical problem to be solved by this invention can also be further achieved through the following technical solution: For the parametric 3D point cloud simulation generation method for circular wall measuring instruments described above, anomaly repair of periodic event point sequences includes statistically analyzing the intervals between adjacent periodic events. ,when When the detection range is exceeded, the missed or false detection events are merged, split, or interpolated for correction.

[0023] The technical problem to be solved by this invention can also be further achieved through the following technical solution: For the parametric three-dimensional point cloud simulation generation method for the circular wall measuring instrument described above, the phase parameters satisfy... , and according to Generate uniform phase sampling points, and obtain a two-dimensional expanded matrix by uniformly resampling the phase of each loop of the signal through interpolation. Where k represents the index of the current sampling point, k n This represents the index of a reference sampling point in the nth round of the signal. It is the uniform phase value corresponding to the j-th column in the two-dimensional expanded matrix, where j is the column index in the two-dimensional expanded matrix.

[0024] The technical problem to be solved by this invention can also be further achieved through the following technical solution: For the parametric three-dimensional point cloud simulation generation method for the above-described circular wall measuring instrument, the phase sampling number W is determined by the number of sampling points per revolution. The statistical value is determined or selected from the candidate set to achieve the optimal number of phase samples W for circumferential closure continuity.

[0025] The technical problem to be solved by the present invention can also be further achieved through the following technical solution: For the parameterized three-dimensional point cloud simulation generation method for the circular wall measuring instrument described above, the periodic boundary processing includes circumferential smoothing or periodic convolution of the neighborhood of j=0 and j=W−1 so that the circumferential closure is continuous without abrupt changes.

[0026] The technical problem to be solved by this invention can also be further achieved through the following technical solution: For the parametric 3D point cloud simulation generation method for a circular wall measuring instrument described above, the quality factor includes at least a frame check factor. Abnormal code factor Interpolation repair factor Filtering residual factor And may further include a closure correction factor. .

[0027] The technical problem to be solved by this invention can also be further achieved through the following technical solution: For the parametric three-dimensional point cloud simulation generation method for a circular wall measuring instrument described above, the point-level uncertainty U satisfies... ,in For quality factor, These are the weighting coefficients. It is a very small positive number used to avoid the denominator being zero. clip is a function that truncates the input to the interval [0,1].

[0028] The technical problem to be solved by the present invention can also be further achieved through the following technical solution: For the parameterized three-dimensional point cloud simulation generation method for the circular wall measuring instrument described above, the axial position parameter z(i) adopts a fixed step Δz or a dynamic step per revolution. z(i) is obtained through cumulative calculation.

[0029] The technical problem to be solved by this invention can also be further achieved through the following technical solution: For the parametric 3D point cloud simulation generation method for a circular wall measuring instrument described above, the radial mapping function is... , where g(·) is a monotonic mapping function.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0031] 1. Resisting rotational speed drift and inconsistent sampling point count: Phase-locked unfolding maps each revolution to a unified phase domain and resamples it to a unified W column, significantly reducing circumferential breakage and point cloud tearing caused by fixed rearrangement misalignment;

[0032] 2. Quality tags can be used for evaluation and robust processing: Point-level U explicitly distinguishes between repair points and reliable points, providing a basis for weighting and elimination for downstream algorithms, and improving the efficiency of system integration and performance evaluation;

[0033] 3. This invention can generate structure-preserving 3D point cloud simulation data without 3D imaging hardware, making it suitable for embedded and engineering scenarios.

[0034] 4. This invention improves the continuity and visualization of the circular wall structure through circumferential closure consistency processing. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the overall process of the method of the present invention;

[0036] Figure 2 This is a schematic diagram of the phase-locked deployment of the present invention;

[0037] Figure 3 This is a schematic diagram illustrating the generation of the point-level uncertainty U according to the present invention. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Reference Figure 1-3 A parametric 3D point cloud simulation generation method for a circular wall measuring instrument is proposed. Specifically, it is a parametric 3D point cloud simulation generation method for a one-dimensional hexadecimal encoded voltage signal of a circular wall measuring instrument. The method involves: analyzing and preprocessing the one-dimensional hexadecimal voltage signal, constructing phase parameters based on periodic event points, and uniformly resampling the phase for each revolution to construct a two-dimensional expansion matrix. Then, parametric 3D mapping is performed to generate point-level uncertainty U, and 3D point cloud simulation data with quality labels is output.

[0040] Among them, the three-dimensional point cloud is a simulated point cloud synthesized based on the one-dimensional signal response characteristics and the prior knowledge of the circular wall structure. It is used for algorithm verification, system integration, visualization and performance evaluation, and is not the same as the point cloud data collected by real three-dimensional imaging equipment.

[0041] The purpose of this method is to address the issue of expansion offset caused by inconsistent sampling points per cycle during actual acquisition. It proposes a phase-locked expansion mechanism based on periodic event points to reduce structural tearing caused by 2D expansion misalignment. It also proposes a point-level uncertainty U generation mechanism to enable the generated point cloud to have quality labeling capabilities, providing a basis for weighting, elimination, and reliability assessment for downstream algorithms.

[0042] The specific scheme of this method is as follows:

[0043] (1) Perform frame verification, code value parsing, and abnormal code marking on the hexadecimal encoded data stream;

[0044] (2) Perform physical quantity restoration, noise reduction, glitch removal, missing segment repair, and normalization on the voltage sequence obtained by analysis;

[0045] (3) Extract the periodic event point sequence from the normalized signal, construct the phase parameter and perform phase uniform resampling on the data of each revolution to generate a two-dimensional expansion matrix, thereby reducing the misalignment caused by the speed drift and the inconsistency of the number of sampling points;

[0046] (4) Apply circumferential closure consistency processing to the two-dimensional expansion matrix and perform parameterized three-dimensional mapping to obtain the three-dimensional point coordinates and intensity attributes;

[0047] (5) During the analysis, repair, filtering and closure process, a quality factor is formed and a point-level uncertainty U is generated by fusion, and the output is point cloud simulation data containing (x,y,z,I,U).

[0048] Specifically:

[0049] I. Symbol and Parameter Definitions

[0050] For ease of description, the present invention uses the following symbols:

[0051] raw_hex: the raw data stream encoded in hexadecimal; c[k]: the integer code value obtained from parsing; V[k]: the voltage sequence; u[k]: the normalized sequence;

[0052] : Periodic event point index sequence; φ(k): phase parameter; W: number of phase sampling columns; H: number of cycles / layers, where the data of a cycle that defines adjacent periodic event points corresponds to a row of a two-dimensional expanded matrix, so the row index i is equivalent to the cycle number n (and can be used interchangeably), and corresponds to the axial layer number, j is the column index; the corresponding axial coordinates From fixed steps or dynamic step Accumulated; M[i,j]: Two-dimensional expansion matrix;

[0053] θ(j): circumferential angle; z(i): axial position; r(i,j): radius; I: strength attribute; U: point-level uncertainty. A larger value indicates that the point is less reliable;

[0054] Interpolation operators are used to perform linear or spline interpolation on discrete sequences within a circle at non-integer indices.

[0055] P: A point cloud set, with points (x, y, z, I, U).

[0056] II. Hexadecimal data reading, frame verification, and exception code marking

[0057] Read the hexadecimal data stream `raw_hex` output from the circular wall measuring instrument and frame it according to the communication protocol. Perform consistency checks on the frame header, frame trailer, length, and checksum field for each frame. Mark frames that fail the check or have missing fields, and record the frame quality factor of the corresponding sampling point as [value missing]. (Or set as the maximum uncertainty), the points that pass the verification are recorded as... .

[0058] Abnormal code detection is performed on the sampled code values, including but not limited to all 0s, all Fs, overflow codes, saturation codes, or code values ​​exceeding the range, and an abnormal code quality factor is generated. Then, based on the rules of bit width, end order, complement, and sign bit, the integer code value sequence c[k] is obtained by parsing.

[0059] III. Voltage Value Restoration and Signal Preprocessing

[0060] Restore the integer code value sequence c[k] to the voltage sequence V[k]:

[0061]

[0062] in For proportionality coefficient, The bias is zero; V[k] is de-scratched and noise suppressed, preferably using sliding window mid-range filtering; missing or damaged segments are repaired by linear interpolation or spline interpolation, and the interpolation repair quality factor is recorded. .

[0063] Normalizing the voltage sequence to [0,1] yields u[k]:

[0064]

[0065] in >0 represents a very small constant, thus avoiding division by zero operations.

[0066] In the implementation method, the filter residual or local noise intensity can be calculated as the quality factor. ,For example And normalize.

[0067] IV. Periodic Event Extraction and Phase Parameter Construction

[0068] To address the issue of misalignment caused by inconsistent sampling points per cycle in actual data acquisition, this invention extracts a sequence of periodic event points from u[k]. , This is used to define the signal interval for each cycle. Periodic event points can be local peaks, zero-crossing points, or correlation peaks with the template signal. To improve robustness, a minimum interval constraint is introduced for event detection. The amplitude threshold is used, and pseudo-events in areas with dense abnormal codes are ignored.

[0069] Interval between adjacent events If statistics are performed, If the value deviates significantly from the median range, it is determined that there may be missed detections / false positives. For example, it can be set as follows: For the circle length sequence The median, when The length of the cycle is determined to be abnormal, among which The preset ratio threshold is used. This invention can repair the event sequence by merging adjacent events, interpolating to fill in points, or removing isolated events, thereby stabilizing each interval.

[0070] Define phase parameters at any sampling point in the nth cycle. Its value range is :

[0071]

[0072] V. Phase-locked expansion and two-dimensional matrix construction

[0073] Set a uniform number of phase sampling columns W. Construct uniform phase sampling points for each revolution. The phase-interpolated resampling of u[k] yields a two-dimensional expanded matrix M[n,j]. This process maps non-uniform sampling sequences of different lengths for each loop into a unified W-column phase domain representation, thereby aligning the circumferential structure.

[0074]

[0075]

[0076]

[0077] in It is a linear interpolation or spline interpolation operator.

[0078] Treating the column direction as a periodic boundary, circumferential smoothing or periodic convolution is applied to the neighborhoods of j=0 and j=W−1 to ensure continuity without abrupt changes at the closure points. In one implementation, the normalized correction amount for the closure process is used as... .

[0079] VI. Parametric 3D Mapping and Point Cloud Generation

[0080] Mapping the two-dimensional unfolded matrix to three-dimensional space, the circumferential angle parameter is:

[0081]

[0082] The axial position parameter can be a fixed step Δz:

[0083]

[0084] Or, when the progress is uneven, use dynamic stepping per revolution. And accumulated The radial mapping is:

[0085]

[0086] in For a monotonic mapping function (which can be linear or piecewise monotonic), the coordinates are transformed as follows:

[0087]

[0088]

[0089]

[0090] This yields the initial set of point clouds, where the intensity attribute I(i,j) can take the value of M[i,j] or the value corresponding to the restoration voltage V.

[0091] VII. Point-level confidence / uncertainty U generation and binding output

[0092] To enable point cloud data to have quality labeling capabilities, this invention generates multiple quality factors during the parsing, repair, filtering, and closure processes, and synchronously maps the quality information to the (i,j) domain along with phase resampling. The quality factors include at least: frame check factor. Abnormal code factor Interpolation repair factor Filtering residual factor and closure correction factor The synchronization mapping preferably uses a two-dimensional expansion matrix. Same phase inverse index and the same interpolation operator This allows us to obtain point-level representations of each quality factor. .

[0093] Point-level uncertainty U can be fused as follows:

[0094]

[0095]

[0096] in The weighting coefficients are denoted by . The final output point cloud is (x, y, z, I, U), which allows downstream algorithms to perform weighted fitting, threshold removal, and reliability assessment based on U.

[0097] 8. Point Cloud Density Equalization, Local Rate of Change Constraints, and Output

[0098] To improve the stability of the spatial distribution of point clouds, voxel statistics and density control can be performed on the point clouds: downsampling of overly dense voxels and supplementing overly sparse voxels; and setting a threshold limit on the radial change rate of the neighborhood to avoid non-physical abrupt changes, and finally outputting point cloud data P with U attribute.

[0099] The advantages of this invention are:

[0100] 1. Phase-locked deployment: Define each lap interval by periodic event points and establish phase parameters. Perform uniform phase resampling for each lap with a uniform phase sampling number W. Overcome the circumferential phase slip caused by rotational speed drift and uneven propulsion under fixed rearrangement, and significantly reduce point cloud diagonal lines, breaks and trailing shadows.

[0101] 2. Point-level uncertainty U: A quality factor is formed in the frame verification, abnormal code, interpolation repair, filtering residual and closure fusion stages and propagated to the point level, outputting a point cloud in the form of (x,y,z,I,U), which provides a basis for weighting, elimination and reliability assessment for downstream algorithms, and improves the efficiency of joint debugging and evaluation.

Claims

1. A parametric 3D point cloud simulation generation method for a circular wall measuring instrument, characterized in that: The steps of this method are as follows: (1) Read the hexadecimal encoded data stream raw_hex output by the circular wall measuring instrument, verify the data frame and parse it to obtain the integer code value sequence. And mark the exception codes; (2) The obtained integer code value sequence Restored to voltage sequence : in For proportionality coefficient, Zero bias; Then, denoising and missing segment repair were performed, and the voltage sequence was normalized to [0,1] to obtain u[k]: in >0 represents a very small constant, thus avoiding division by zero operations; (3) In Extracting periodic event point sequences This is used to define the signal interval for each revolution and to construct phase parameters within each interval. ; (4) Perform uniform phase resampling on each loop of signal using a uniform number of phase samples W to construct a two-dimensional expansion matrix. Periodic boundary treatment is applied to the column direction to ensure circumferential closure consistency, and uniform phase sampling points are constructed for each revolution. ; ； ； in For linear interpolation or spline interpolation operators; (5) Expand the two-dimensional matrix According to circumferential angle parameters Axial position parameters With radial mapping function Perform parametric 3D mapping to generate point coordinates (x, y, z) and intensity attribute I; (6) In the process of frame verification, abnormal code marking, missing segment repair, filtering and circumferential closure processing, a quality factor is formed and fused to generate a point-level uncertainty attribute U, and finally the output is three-dimensional point cloud simulation data containing (x,y,z,I,U).

2. The parametric 3D point cloud simulation generation method for a circular wall measuring instrument according to claim 1, characterized in that: The periodic event points are obtained through local peak detection, zero-crossing detection, or detection related to the template signal, and minimum interval constraints and amplitude thresholds are set during the detection process to suppress glitch events.

3. The parametric 3D point cloud simulation generation method for a circular wall measuring instrument according to claim 1, characterized in that: Anomaly repair for periodic event point sequences includes statistical analysis of the intervals between adjacent periodic events. ,when When the detection range is exceeded, the missed or false detection events are merged, split, or interpolated for correction.

4. The parametric 3D point cloud simulation generation method for a circular wall measuring instrument according to claim 1, characterized in that: The phase parameters satisfy , and according to Generate uniform phase sampling points, and obtain a two-dimensional expanded matrix by uniformly resampling the phase of each loop of the signal through interpolation. Where k represents the index of the current sampling point, k n This represents the index of a reference sampling point in the nth round of the signal. It is the uniform phase value corresponding to the j-th column in the two-dimensional expanded matrix, where j is the column index in the two-dimensional expanded matrix.

5. The parametric 3D point cloud simulation generation method for a circular wall measuring instrument according to claim 1, characterized in that: The number of phase samples W is determined by the number of sampling points per revolution. The statistical value is determined or selected from the candidate set to achieve the optimal number of phase samples W for circumferential closure continuity.

6. The method according to claim 1, characterized in that, The periodic boundary processing includes circumferential smoothing or periodic convolution of the neighborhoods of j=0 and j=W−1 to ensure that the circumferential closure is continuous without abrupt changes.

7. The parametric 3D point cloud simulation generation method for a circular wall measuring instrument according to claim 1, characterized in that: The quality factor includes at least a frame check factor. Abnormal code factor Interpolation repair factor Filtering residual factor And may further include a closure correction factor. .

8. The parametric three-dimensional point cloud simulation generation method for a circular wall measuring instrument according to claim 1 or 7, characterized in that: The point-level uncertainty U satisfies ,in For quality factor, These are the weighting coefficients. It is a very small positive number used to avoid the denominator being zero. clip is a function that truncates the input to the interval [0,1].

9. The parametric 3D point cloud simulation generation method for a circular wall measuring instrument according to claim 1, characterized in that: The axial position parameter z(i) is obtained by either a fixed step Δz or a dynamic step per revolution. z(i) is obtained by cumulative calculation.

10. The parametric three-dimensional point cloud simulation generation method for a circular wall measuring instrument according to claim 1, characterized in that: The radial mapping function is: , where g(·) is a monotonic mapping function.

Images