A displacement measurement method and related device using optical topological state

Abstract

The present application relates to the field of displacement measurement technology, and discloses a displacement measurement method and related devices using optical topological states. The method comprises: irradiating a target object with a preset laser emitting device, collecting light signals with an optical sensor, and converting the light signals into digital signals; removing background noise from the digital signals to obtain clean digital signals; identifying topological feature points of the clean digital signals to obtain a set of topological feature points; calculating the displacement vector of the target object based on the set of topological feature points to obtain a target displacement vector. The present application uses the optical topological state to measure the accuracy of displacement.

Description

Technical Field

[0001] The present application relates to the field of displacement measurement, and in particular to a displacement measurement method and related devices using optical topological states. Background Art

[0002] In today's high-precision measurement needs, especially in industrial inspection, machine vision, and scientific research, higher requirements are placed on the accuracy and efficiency of displacement measurement. Traditional displacement measurement methods, such as mechanical contact measurement and basic optical measurement, are still effective in specific applications, but they face many challenges when dealing with complex environments or when non-contact measurement is required. Mechanical contact measurement may affect the state of the measured object due to physical contact, while basic optical measurement methods often cannot meet more stringent scientific research and industrial standards in terms of accuracy and stability. These methods are particularly limited when dealing with small displacements, fast-moving objects, or measurements in unstable environments.

[0003] In recent years, with the rapid development of optical technology and image processing technology, the method of displacement measurement using optical topological states has attracted widespread attention. This method calculates the displacement of the target object by analyzing the optical topological changes on the surface of the target object captured by the optical sensor. Although this method provides the possibility of non-contact measurement and is expected to greatly improve the accuracy and adaptability of measurement, there are still some challenges in practical applications. For example, how to accurately extract topological state feature points from complex background noise and how to deal with changes in optical properties caused by environmental changes are key issues that need to be solved. Summary of the invention

[0004] The present application provides a displacement measurement method and related devices using optical topological states, which are used to improve the accuracy of displacement measurement using optical topological states.

[0005] In a first aspect, the present application provides a displacement measurement method using an optical topological state, the displacement measurement method using an optical topological state comprising: irradiating a target object with a preset laser emitting device, collecting an optical signal with an optical sensor, and converting the optical signal into a digital signal;

[0006] removing background noise from the digital signal to obtain a clean digital signal;

[0007] Performing topological feature point recognition on the clean digital signal to obtain a topological feature point set;

[0008] Based on the topological feature point set, a displacement vector of the target object is calculated to obtain a target displacement vector.

[0009] In combination with the first aspect, in a first implementation of the first aspect of the present application, irradiating the target object by a preset laser emitting device, collecting light signals by an optical sensor, and converting the light signals into digital signals include:

[0010] The target object is irradiated by a preset laser emitting device, and at the same time, the object attributes of the target object are extracted to obtain the target object attributes;

[0011] Based on the properties of the target object, collecting light reflection characteristic information of the target object;

[0012] configuring the sampling frequency of the optical sensor according to the light reflection characteristic information, and collecting the light signal through the optical sensor;

[0013] Calculating the average power of the optical signal to obtain the average power of the signal, and inputting the average power of the signal into a preset automatic gain control algorithm to perform input gain correction to obtain a target input gain;

[0014] The target input gain is input into a preset analog signal converter for digital signal conversion to obtain the digital signal.

[0015] In combination with the first aspect, in a second implementation of the first aspect of the present application, the average power of the optical signal is calculated to obtain the signal average power, and the signal average power is input into a preset automatic gain control algorithm for input gain correction to obtain the target input gain, including:

[0016] The instantaneous power of the optical signal is calculated to obtain an instantaneous power set, wherein the expression of the instantaneous power is as follows:

[0017] ;in, is the signal amplitude at time t, is the instantaneous power;

[0018] The instantaneous power set is subjected to sliding average processing to obtain the signal average power, wherein the sliding average calculation formula is as follows:

[0019] ; Where M is the time window length, is the average signal power, i is the time index value;

[0020] Based on the signal average power, the signal power standard deviation of the optical signal is calculated to obtain the signal power standard deviation, wherein the signal power standard deviation calculation formula is as follows:

[0021] ;in, is the signal power standard deviation;

[0022] Based on the signal power standard deviation and the signal average power, a gain adjustment factor is calculated for the optical signal using a gain adjustment factor calculation formula to obtain a gain adjustment factor value, wherein the gain adjustment factor calculation formula is as follows:

[0023] ;in, is the gain adjustment factor value, is the preset target power data, is the adjustment coefficient;

[0024] A gain adjustment value is calculated based on the gain adjustment factor value to obtain a target gain adjustment value, wherein the gain adjustment value calculation formula is as follows:

[0025] ;in, is the target gain adjustment value, is the conversion factor value;

[0026] An input gain correction is performed based on the target gain adjustment value to obtain a target input gain.

[0027] In combination with the first aspect, in a third implementation manner of the first aspect of the present application, removing background noise from the digital signal to obtain a clean digital signal includes:

[0028] Performing spectrum recognition on the digital signal to obtain a background noise frequency band;

[0029] Based on the background noise frequency band matched filter, a plurality of initial filters are obtained;

[0030] Calculating the cutoff frequency of each of the initial filters to obtain a cutoff frequency set;

[0031] Based on the cutoff frequency set, parameter configuration is performed on each of the initial filters to obtain a plurality of target filters;

[0032] Decomposing the digital signal to obtain multiple sub-signals;

[0033] Performing signal recombination on the plurality of sub-signals to obtain a recombined signal;

[0034] The background noise of the recombined signal is removed based on the multiple target filters to obtain the clean digital signal.

[0035] In combination with the first aspect, in a fourth implementation of the first aspect of the present application, the step of performing topological feature point identification on the clean digital signal to obtain a topological feature point set includes:

[0036] Performing image reconstruction on the clean digital signal to obtain reconstructed image data;

[0037] Segmenting the reconstructed image data into key parts to obtain a plurality of key part data;

[0038] Based on each of the key part data, locally enhance the contrast of the reconstructed image to obtain enhanced image data;

[0039] Performing feature point position recognition on the enhanced image data to obtain a feature point position set;

[0040] Performing image segmentation on the enhanced image data based on the feature point position set to obtain a plurality of initial feature point images;

[0041] Performing image screening on the multiple initial feature point images to obtain multiple target feature point images;

[0042] Topological feature point recognition is performed on a plurality of the target feature point images to obtain a topological feature point set.

[0043] In combination with the first aspect, in a fifth implementation of the first aspect of the present application, the step of performing topological feature point recognition on the plurality of target feature point images to obtain a topological feature point set includes:

[0044] Constructing a feature point descriptor for each of the target feature point images to obtain a descriptor set;

[0045] Based on the descriptor set, performing image pair matching on the plurality of target feature point images by a preset matching algorithm to obtain a plurality of groups of image pairs;

[0046] Performing initial screening of topological feature points on the plurality of groups of image pairs to obtain an initial feature point set;

[0047] The initial feature point set is deduplicated to obtain the topological feature point set.

[0048] In combination with the first aspect, in a sixth implementation of the first aspect of the present application, the step of calculating a displacement vector of the target object based on the set of topological feature points to obtain a target displacement vector includes:

[0049] Performing cross-frame matching on the target object to obtain multiple consecutive frame images;

[0050] Using the topological feature point set, image displacement point recognition is performed on multiple frames of the continuous frame images to obtain multiple groups of displacement points;

[0051] Based on the multiple groups of displacement points, a displacement vector of the target object is calculated to obtain a target displacement vector.

[0052] In a second aspect, the present application provides a displacement measurement device using an optical topological state, the displacement measurement device using an optical topological state comprising:

[0053] The acquisition module is used to illuminate the target object through a preset laser emitting device, collect light signals through an optical sensor, and convert the light signals into digital signals;

[0054] A removal module, used for removing background noise from the digital signal to obtain a clean digital signal;

[0055] An identification module, used for performing topological feature point identification on the clean digital signal to obtain a topological feature point set;

[0056] The calculation module is used to calculate the displacement vector of the target object based on the topological feature point set to obtain the target displacement vector.

[0057] A third aspect of the present application provides a displacement measurement device utilizing optical topological states, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the displacement measurement device utilizing optical topological states executes the above-mentioned displacement measurement method utilizing optical topological states.

[0058] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned displacement measurement method using optical topological states.

[0059] In the technical solution provided by the present application, the target object is accurately illuminated by a laser emitting device, and the reflected light signal is accurately collected by an optical sensor. This method can capture slight changes on the surface of the target object. The step of removing background noise from the digital signal ensures that the accuracy of the measurement will not be disturbed by environmental noise. Through advanced signal processing technology, the target signal and the background noise can be effectively separated to obtain a clean digital signal, which provides a reliable basis for subsequent feature point recognition and displacement vector calculation. Using image processing and computer vision technology to identify topological feature points of clean digital signals can accurately determine the key feature points on the surface of the target object. By analyzing the set of topological feature points, the method can accurately calculate the displacement vector of the target object. This not only improves the accuracy of displacement measurement, but also greatly improves the calculation efficiency by focusing on key feature points. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0061] Figure 1 This is a schematic diagram of an embodiment of a displacement measurement method using an optical topological state in an embodiment of the present application;

[0062] Figure 2 Schematic diagram of an embodiment of a displacement measurement device using optical topological states in an embodiment of the present application. DETAILED DESCRIPTION

[0063] The present application embodiment provides a displacement measurement method and related device using optical topological states. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0064] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiments of the present application, an embodiment of the displacement measurement method using optical topological states includes:

[0065] Step S101, irradiating a target object with a preset laser emitting device, collecting light signals with an optical sensor, and converting the light signals into digital signals;

[0066] It is understandable that the execution subject of the present application may be a displacement measurement device using an optical topological state, or may be a terminal or a server, which is not specifically limited here. The present application embodiment is described by taking a server as the execution subject as an example.

[0067] Specifically, the target object is illuminated by a preset laser emitting device, and at the same time, the object attributes of the target object are extracted to obtain the key attribute information of the target object. The attributes of the target object, such as material, color and surface structure, will directly affect its light reflection characteristics. Based on the attributes of the target object, the light reflection characteristic information of the target object is collected, which helps to understand how light is reflected by the target object and received by the optical sensor. The optical sensor is configured with the collected light reflection characteristic information to ensure that the sensor collects the reflected light signal at the most suitable frequency to maximize the signal quality and reduce noise. The collected light signal is then averaged to obtain the average power of the signal. The calculated average power is an important indicator for evaluating the overall strength of the signal, reflecting the average energy level of the light signal reflected from the target object. After obtaining the average power, the value is input into the preset automatic gain control algorithm. The automatic gain control algorithm automatically adjusts the input gain according to the average power of the signal to optimize the dynamic range and signal-to-noise ratio of the signal and ensure the quality of the signal during the digital signal conversion process. The target input gain is input into a preset analog signal converter for digital signal conversion. The analog signal converter converts the gain-corrected analog optical signal into a digital signal, completing the conversion process from physical optical signal to digital information.

[0068] Among them, the instantaneous power of the optical signal is calculated to obtain the instantaneous power set. The instantaneous power is obtained by the square of the signal amplitude, which reflects the energy of the signal at any given moment. The instantaneous power is processed by sliding average, and the fluctuation of the instantaneous power is smoothed by calculating the average power within a predetermined time window to obtain a more stable and reliable signal average power. The calculation of the sliding average not only reduces the impact of noise, but also provides a clear benchmark for further processing of the signal. The standard deviation of the signal power is calculated based on the signal average power, and the fluctuation range of the signal power is evaluated by measuring the deviation between the power value and its average value. The standard deviation of the signal power is an important parameter for evaluating signal stability and directly affects the formulation of the gain adjustment strategy. The gain adjustment factor is calculated by combining the average power and the power standard deviation of the signal through a specific gain adjustment factor calculation formula. The gain adjustment factor takes into account the target power data, the average power of the signal and its volatility, and balances the signal stability and gain adjustment requirements through the adjustment coefficient λ. The calculation of the gain adjustment factor is the key to the entire gain adjustment process, which directly determines the direction and amplitude of the gain adjustment. Based on the gain adjustment factor value, the target gain adjustment value is calculated, and the final gain adjustment value is obtained by logarithmically converting the gain adjustment factor and multiplying it by the conversion factor. This gain adjustment value will be used to correct the input gain to ensure that the signal has an appropriate gain level before entering the digital processing, and will not cause saturation due to excessive signal strength, nor lose details due to excessive signal weakness.

[0069] Step S102, removing background noise from the digital signal to obtain a clean digital signal;

[0070] It should be noted that spectrum recognition is performed on digital signals to determine which frequency bands contain background noise. Spectrum recognition identifies which frequency bands are mainly composed of noise by analyzing the frequency domain characteristics of the signal. Based on the identified background noise bands, multiple initial filters are obtained. These filters may have different types and parameters in the initial stage to cover and eliminate the identified noise bands as much as possible. In order to optimize these filters, the cutoff frequency is calculated for each initial filter to obtain a cutoff frequency set. The set reflects the frequency boundary that needs to be set to effectively filter out background noise. The determination of the cutoff frequency directly affects the performance of the filter, including its retention of the signal and the removal of noise. Based on the cutoff frequency set, the parameters of each initial filter are configured to obtain multiple target filters. After adjustment, the target filter can more effectively remove the background noise while retaining the integrity and quality of the target signal as much as possible. The parameter configuration needs to comprehensively consider the type of filter, the cutoff frequency, and the characteristics of the signal itself. In order to further improve the effect of noise removal, the digital signal is decomposed to obtain multiple sub-signals. The purpose of signal decomposition is to decompose a complex signal into several parts that are easier to process, and each of these sub-signals contains different components of the signal. These sub-signals are recombined to form a recombined signal. The recombined signal inherits the basic characteristics of the original signal, but through the process of decomposition and recombination, the noise components in the signal are more clearly exposed, which is convenient for subsequent denoising. The background noise of the recombined signal is removed based on multiple target filters. The target filters will be applied to the recombined signal in sequence or in combination to remove the noise components and finally obtain a clean digital signal.

[0071] Step S103, performing topological feature point recognition on the clean digital signal to obtain a topological feature point set;

[0072] The clean digital signal is reconstructed to convert the digital signal into image data. The reconstructed image data is segmented for key parts to distinguish the parts containing key information from other parts of the image and obtain multiple key parts data. During the segmentation process, the key parts in the image are identified and extracted by applying image processing algorithms such as edge detection, region growing or threshold processing. Based on each key part data, the reconstructed image is locally contrast enhanced to make the features in the image more obvious and facilitate the subsequent feature point recognition. Local contrast enhancement improves the image quality by adjusting the brightness and contrast of specific areas in the image, making the feature points easier to identify. Feature point position recognition is performed on the enhanced image data. By analyzing the local features of the image, such as corners, edges or textures, potential feature point positions are identified to form a feature point position set. This set represents the position information of the topological feature points in the image. Based on the feature point position set, the enhanced image data is segmented to separate the area containing potential feature points from the image to obtain multiple initial feature point images. These initial feature point images collectively represent all possible feature points, but may contain some irrelevant or misidentified parts. Perform image screening, and screen out those images that are most likely to represent real topological feature points by evaluating the quality and relevance of the initial feature point images, and obtain multiple target feature point images. Ensure the accuracy and reliability of the final identified feature point set by removing noise and irrelevant image content. Perform topological feature point recognition on the target feature point image, and use image processing and analysis technology to identify the real topological feature points from the target feature point image and form a topological feature point set. This set contains all the identified feature points and represents the displacement information of the target object.

[0073] Among them, a feature point descriptor is constructed for each target feature point image, and the feature points in the image are converted into a mathematical model that can be understood and processed by a computer, providing a quantifiable and comparable method for matching between images. These descriptor sets contain sufficient information to reflect the uniqueness of each feature point and the information of its surrounding area. Based on the constructed descriptor set, multiple target feature point images are matched by a preset matching algorithm. The preset matching algorithm may include nearest neighbor matching, sequence minimum optimization matching or other efficient image matching techniques, aiming to find feature points with the same descriptors between different images to form multiple groups of image pairs. The formation of image pairs is based on the similarity of feature points in the image, which provides a basis for identifying stable and consistent topological feature points. The topological feature points of the formed multiple groups of image pairs are initially screened, and those feature points that appear repeatedly and match consistently in different images are identified from the multiple groups of image pairs to obtain the initial feature point set. The points that are most likely to represent the topological features of the object are screened out, and those points that are accidental or mismatched are removed. The initial feature point set is deduplicated to ensure that the feature point set does not contain duplicate feature points, and a topological feature point set is obtained.

[0074] Step S104: Based on the topological feature point set, a displacement vector of the target object is calculated to obtain a target displacement vector.

[0075] Specifically, the target object is matched across frames to obtain multiple continuous frame images. Cross-frame matching identifies the corresponding positions of the same target object in different images by comparing images captured at different time points. Image displacement points are identified for multiple continuous frame images through a set of topological feature points. The stability and uniqueness of the topological feature points are used to identify the displacement points representing the position change of the target object in the continuous frame images. By analyzing the position changes of these displacement points in different frame images, the movement trajectory of the target object over time is tracked. Image displacement point recognition is based on the analysis of the position change of each feature point between continuous frames. In this way, multiple groups of displacement points are obtained, each group of displacement points represents the position change of the target object between two continuous time points. Based on multiple groups of displacement points, the displacement vector of the target object is calculated. The relative position change between the displacement points is converted into a mathematical vector expression to obtain a target displacement vector. Considering the position change of the displacement point on the image plane and the change in the time dimension, it is ensured that the displacement vector can accurately reflect the actual movement of the target object. By comprehensively analyzing multiple groups of displacement points, the calculated target displacement vector contains information about the moving direction and the moving distance.

[0076] In the embodiment of the present application, the target object is accurately illuminated by a laser emitting device, and the reflected light signal is accurately collected by an optical sensor. This method can capture slight changes on the surface of the target object. The step of removing background noise from the digital signal ensures that the accuracy of the measurement will not be disturbed by environmental noise. Through advanced signal processing technology, the target signal and the background noise can be effectively separated to obtain a clean digital signal, which provides a reliable basis for subsequent feature point recognition and displacement vector calculation. Using image processing and computer vision technology to identify topological feature points of clean digital signals can accurately determine the key feature points on the surface of the target object. By analyzing the set of topological feature points, the method can accurately calculate the displacement vector of the target object. This not only improves the accuracy of displacement measurement, but also greatly improves the calculation efficiency by focusing on key feature points.

[0077] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0078] (1) Irradiating the target object through a preset laser emitting device, and at the same time, extracting the object attributes of the target object to obtain the target object attributes;

[0079] (2) Based on the properties of the target object, collect the light reflection characteristic information of the target object;

[0080] (3) configuring the sampling frequency of the optical sensor based on the light reflection characteristic information, and collecting the light signal through the optical sensor;

[0081] (4) Calculate the average power of the optical signal to obtain the average power of the signal, and input the average power of the signal into a preset automatic gain control algorithm to perform input gain correction to obtain the target input gain;

[0082] (5) The target input gain is input into a preset analog signal converter for digital signal conversion to obtain a digital signal.

[0083] Specifically, the target object is irradiated by a preset laser emitting device, and at the same time, the object attributes of the target object are extracted, including the recognition of the surface characteristics, material type, color and other related attributes of the target object. For example, when the laser is irradiated on a metal ball, the smoothness of its surface, the type of metal and the color of the ball will affect the reflection characteristics of the laser light. Based on the attributes of the target object, the light reflection characteristic information of the target object is collected. For example, due to its smooth surface and material characteristics, the metal ball may cause obvious laser reflection and form a strong light spot; while the cloth may have weak laser reflection and stronger dispersion due to its surface roughness and light absorption. After obtaining the light reflection characteristic information, the sampling frequency of the optical sensor is configured according to this information. According to the intensity and distribution characteristics of the reflected light, the sampling frequency of the sensor is adjusted to best capture the reflected light signal. For example, for the strong reflection on the surface of the metal ball, a lower sampling frequency may be required to avoid signal saturation; while for the weak reflection on the cloth, a higher sampling frequency may be required to ensure the effective capture of the signal. The average power of the collected light signal is calculated, and the overall strength of the signal is evaluated by analyzing the power distribution of the signal. The calculated average power of the signal is an important indicator for evaluating signal quality. It can help the system determine whether the signal is within a reasonable range and whether it needs to be optimized through gain adjustment. For example, if the average power is too high, it may indicate that the signal is close to saturation and the gain needs to be reduced; if the average power is too low, the gain may need to be increased to improve the signal quality. The signal average power is input into the preset automatic gain control algorithm for input gain correction. The automatic gain control algorithm automatically adjusts the input gain according to the signal average power to optimize the dynamic range and signal-to-noise ratio of the signal. For example, for a strong light signal reflected from a metal ball, the algorithm may reduce the gain to avoid signal oversaturation; for a weak light signal reflected from a cloth, the algorithm may increase the gain to ensure the detectability of the signal. The adjusted target input gain is input into the preset analog signal converter for digital signal conversion. Convert the analog optical signal to a digital format for subsequent processing and analysis.

[0084] In a specific embodiment, the process of performing the step of calculating the average power of the optical signal may specifically include the following steps:

[0085] (1) Calculate the instantaneous power of the optical signal to obtain the instantaneous power set, where the expression of instantaneous power is as follows:

[0086] ;in, is the signal amplitude at time t, is the instantaneous power;

[0087] (2) Perform sliding average processing on the instantaneous power set to obtain the signal average power, where the sliding average calculation formula is as follows:

[0088] ; Where M is the time window length, is the average signal power, i is the time index value;

[0089] (3) Based on the signal average power, the signal power standard deviation of the optical signal is calculated to obtain the signal power standard deviation, where the signal power standard deviation calculation formula is as follows:

[0090] ;in, is the signal power standard deviation;

[0091] (4) Based on the signal power standard deviation and the signal average power, the gain adjustment factor of the optical signal is calculated using the gain adjustment factor calculation formula to obtain the gain adjustment factor value, where the gain adjustment factor calculation formula is as follows:

[0092] ;in, is the gain adjustment factor value, is the preset target power data, is the adjustment coefficient;

[0093] (5) Calculate the gain adjustment value based on the gain adjustment factor value to obtain the target gain adjustment value, where the gain adjustment value calculation formula is as follows:

[0094] ;in, is the target gain adjustment value, is the conversion factor value;

[0095] (6) Based on the target gain adjustment value, the input gain is corrected to obtain the target input gain.

[0096] Specifically, the instantaneous power of the optical signal is calculated, and the instantaneous power set is obtained based on the square of the signal amplitude, and each element reflects the power level of the signal at a specific moment. The average power of the signal is obtained by performing sliding average processing on the instantaneous power set. Sliding average is a common signal processing technique used to smooth time series data. By calculating the average value within a certain time window to reduce the impact of instantaneous changes, a more stable and reliable average power estimate is obtained. Based on the obtained signal average power, the signal power standard deviation is calculated. The variability of the signal power is analyzed, that is, the degree of deviation of the power value at different time points relative to the average power. Calculating the signal power standard deviation helps to understand the stability of the signal. For example, if the signal power standard deviation is large, it indicates that the signal power changes greatly, and a more detailed gain adjustment strategy may be required to ensure stable reception of the signal. Based on the signal average power and the signal power standard deviation, the gain adjustment factor value is calculated through a specific gain adjustment factor calculation formula. The automatic gain control algorithm determines how to adjust the input gain of the system in order to achieve the preset target power level. For example, if the target power is set to a specific value and the current average power is lower than this value, the gain adjustment factor will instruct the system to increase the input gain. The adjustment factor adjusts the gain adjustment factor according to the variability of the power to ensure that the gain adjustment can achieve the target power without introducing additional noise or distortion due to over-adjustment. The gain adjustment factor is converted into a specific gain adjustment value, and the value of the gain adjustment factor is converted into an actual gain setting parameter. For example, by taking the logarithm of the gain adjustment factor and multiplying it by a conversion factor , get the specific gain adjustment value, which is directly used to adjust the input gain of the system.

[0097] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0098] (1) Perform spectrum recognition on the digital signal to obtain the background noise frequency band;

[0099] (2) Based on the background noise band matched filter, multiple initial filters are obtained;

[0100] (3) Calculate the cutoff frequency of each initial filter to obtain a set of cutoff frequencies;

[0101] (4) Based on the cutoff frequency set, configure the parameters of each initial filter to obtain multiple target filters;

[0102] (5) Decompose the digital signal to obtain multiple sub-signals;

[0103] (6) Recombining multiple sub-signals to obtain a recombined signal;

[0104] (7) The background noise of the reconstructed signal is removed based on multiple target filters to obtain a clean digital signal.

[0105] Specifically, spectrum recognition is performed on the digital signal to identify the frequency band occupied by the background noise in the frequency domain of the signal. Spectral recognition converts the signal in the time domain into a representation in the frequency domain by performing Fourier transform (FFT) on the signal. For example, the noise distribution in the spectrum may appear as a part of the spectrum or a peak of a specific frequency. The corresponding initial filter is matched based on the background noise band. For the identified noise bands, the noise in these bands is effectively suppressed or eliminated. For example, if the identified noise is mainly concentrated in the low-frequency area, a high-pass filter may be selected as one of the initial filters to weaken or eliminate the noise in this part of the frequency. The cutoff frequency is calculated for each initial filter. The determination of the cutoff frequency is based on the analysis of the background noise band, aiming to minimize the impact on the useful signal while removing as much noise as possible. For example, by analyzing the boundary of the noise band, the optimal cutoff frequency of the high-pass or low-pass filter is determined, a cutoff frequency set is formed, and one or more operating parameters are defined for each initial filter. Based on the cutoff frequency set, each initial filter is parameterized, and each initial filter is optimized and converted into a target filter by adjusting these parameters. Ensure that each filter provides the best noise suppression effect in a specific application scenario. For example, for a high-pass filter, fine-tuning its cutoff frequency can ensure that low-frequency noise is effectively reduced while the high-frequency components of the useful signal are retained. Perform signal decomposition on the digital signal to further improve the noise removal effect. By decomposing the signal into multiple sub-signals, each sub-signal can be analyzed and processed independently. For example, the sub-signals obtained by wavelet transform decomposition can be processed separately for noise in different frequency bands. Perform signal recombining on multiple sub-signals. Recombine the processed multiple sub-signals into a whole. For example, after the sub-signals processed by wavelet transform have removed the noise in their respective frequency bands, they are recombined through inverse transform to form a clearer and cleaner signal. Use multiple optimized target filters to remove background noise from the recombined signal. Each target filter is optimized for specific noise characteristics, so that background noise can be removed more effectively, and finally a clean digital signal is obtained.

[0106] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0107] (1) Reconstructing the clean digital signal to obtain reconstructed image data;

[0108] (2) Segment the reconstructed image data into key parts to obtain multiple key part data;

[0109] (3) Based on the data of each key part, the reconstructed image is locally contrast enhanced to obtain enhanced image data;

[0110] (4) Identify the feature point positions of the enhanced image data to obtain a feature point position set;

[0111] (5) Performing image segmentation on the enhanced image data based on the feature point position set to obtain multiple initial feature point images;

[0112] (6) Screening multiple initial feature point images to obtain multiple target feature point images;

[0113] (7) Perform topological feature point recognition on multiple target feature point images to obtain a topological feature point set.

[0114] Specifically, the clean digital signal is reconstructed to convert the digital signal into image data so that the information in the signal can be visualized and further analyzed in the form of an image. For example, by applying an appropriate image reconstruction algorithm, such as a method based on Fourier transform, these digital signals are converted into images, in which the surface features and texture patterns of the object are converted into brightness and color changes of the image. The reconstructed image data is segmented for key parts to identify and extract important areas in the image. By using image segmentation techniques such as threshold processing, region growing or edge detection, the key parts in the image are separated from the background or irrelevant areas. For example, the key parts may be specific textures or marks on the surface, which are segmented in the reconstructed image to form multiple key part data. Based on each key part data, the reconstructed image is locally contrast enhanced to improve the visibility and recognition of specific areas in the image. By adjusting the local contrast of the image, the key features are made more obvious, which facilitates further feature point recognition. Contrast enhancement can be achieved through various image processing algorithms, such as histogram equalization or local contrast enhancement technology. Feature point position recognition is performed on the enhanced image data to identify the positions of points that may represent object features from the enhanced image. Feature point location recognition usually relies on image processing techniques such as corner detection, template matching or machine learning methods to identify significant points in an image. For example, in an enhanced surface image of an object, by applying a corner detection algorithm, points with significant brightness changes due to texture changes or other surface features can be identified, and the collection of these points forms a feature point location set. Image segmentation is performed on the enhanced image data, and the image is segmented into multiple image segments containing initial feature points according to the locations of the feature points. A small area is defined around each feature point and these areas are extracted as separate images. Image screening is performed on multiple initial feature point images to identify the image that is most likely to represent the topological state features of the target object. The quality, clarity and information content of the initial feature point images are evaluated, and those images with low quality or that do not match the target features are eliminated, and the most relevant and information-rich images are retained as target feature point images. Topological state feature point recognition is performed on multiple target feature point images, and feature points that truly represent the topological state of the object are identified by analyzing the geometric and topological information in these images. This may involve complex image analysis techniques, including shape-based matching, topological analysis, or deep learning methods to accurately extract a set of topological feature points from the target feature point image.

[0115] In a specific embodiment, the process of performing the topological feature point recognition step on multiple target feature point images may specifically include the following steps:

[0116] (1) Construct a feature point descriptor for each target feature point image to obtain a descriptor set;

[0117] (2) Based on the descriptor set, image pairs are matched on multiple target feature point images using a preset matching algorithm to obtain multiple groups of image pairs;

[0118] (3) Performing initial screening of topological feature points on multiple sets of image pairs to obtain an initial set of feature points;

[0119] (4) De-duplicate the initial feature point set to obtain a topological feature point set.

[0120] Specifically, a feature point descriptor is constructed for each target feature point image. Feature point descriptors are used to describe the local appearance and shape of feature points in an image in a standardized way. These descriptors are not only invariant to the scaling, rotation and brightness changes of the image, but also need to have sufficient discrimination to distinguish them from other feature points in the image. For example, for a set of images of the surface of an object taken from different angles, the feature points in each image (such as concave and convex points or color changes on the surface of the object) can be constructed with descriptors using algorithms such as SIFT (Scale Invariant Feature Transform) or SURF (Speeded Up Robust Features), which can generate a vector containing key local image feature information to form a descriptor set. Based on the constructed descriptor set, multiple target feature point images are matched by a preset matching algorithm to find the same feature points between two or more images. The nearest neighbors are searched in the descriptor space, or more complex matching strategies such as graph-based matching or machine learning methods are used to identify matching pairs. Taking the surface image of an object as an example, the corresponding feature points in different images are identified by the matching algorithm to obtain multiple sets of image pairs. The multiple sets of image pairs are initially screened for topological feature points, and those feature points that truly represent the changes in the topological state of the object are identified from the matching pairs to form an initial set of feature points. The initial screening takes into account the geometric consistency and topological relationship between the feature points to ensure that the selected feature points remain consistent in different states or different viewing angles of the object. For example, when analyzing a bent and deformed object, the initial screening will identify those feature points that can be matched before and after the object is bent. These points reflect the key information of the object's topological structure. The initial set of feature points is deduplicated. In the matching process of multiple sets of image pairs, there may be repeated feature points. The deduplication process is intended to ensure that each feature point is unique in the final set of topological feature points. The set of topological feature points is obtained by comparing the position, descriptors or other attributes of the feature points to identify and remove duplicates.

[0121] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0122] (1) Perform cross-frame matching on the target object to obtain multiple consecutive frame images;

[0123] (2) Using the topological feature point set, image displacement points are identified for multiple consecutive frame images to obtain multiple groups of displacement points;

[0124] (3) Based on multiple sets of displacement points, the displacement vector of the target object is calculated to obtain the target displacement vector.

[0125] Specifically, the target object is matched across frames to identify and track the position changes of the same target object in multiple time series images. Each frame of the image is analyzed to determine the corresponding position of the target object in different frames. Cross-frame matching algorithms, such as feature matching-based algorithms (SIFT, SURF) or deep learning-based target tracking algorithms, are used to identify the position of the target object in the image sequence and track the changes in this position in consecutive frames to obtain multiple consecutive frame images. Image displacement points are identified for multiple consecutive frame images through a set of topological feature points. The algorithm needs to identify feature points representing the topological state of the target object in each frame of the image and track the movement of these feature points between consecutive frames. Based on multiple sets of displacement points, the displacement vector of the target object is calculated to obtain the target displacement vector. The calculation of the displacement vector requires converting the relative position changes between the displacement points into geometric vectors, which represent the movement direction and distance of the target object or its part in space.

[0126] The above describes the displacement measurement method using the optical topological state in the embodiment of the present application. The following describes the displacement measurement device using the optical topological state in the embodiment of the present application. Figure 2 In the embodiments of the present application, an embodiment of a displacement measurement device using an optical topological state includes:

[0127] The acquisition module 201 is used to illuminate the target object through a preset laser emitting device, collect light signals through an optical sensor, and convert the light signals into digital signals;

[0128] A removal module 202, configured to remove background noise from the digital signal to obtain a clean digital signal;

[0129] The recognition module 203 is used to perform topological feature point recognition on the clean digital signal to obtain a topological feature point set;

[0130] The calculation module 204 is used to calculate the displacement vector of the target object based on the topological feature point set to obtain a target displacement vector.

[0131] Through the synergy of the above components, the target object is accurately illuminated by the laser emitting device, and the reflected light signal is accurately collected by the optical sensor. This method can capture the slight changes on the surface of the target object. The step of removing the background noise of the digital signal ensures that the accuracy of the measurement will not be disturbed by the environmental noise. Through advanced signal processing technology, the target signal and the background noise can be effectively separated to obtain a clean digital signal, which provides a reliable basis for the subsequent feature point recognition and displacement vector calculation. The topological feature point recognition of the clean digital signal using image processing and computer vision technology can accurately determine the key feature points on the surface of the target object. By analyzing the set of topological feature points, this method can accurately calculate the displacement vector of the target object. This not only improves the accuracy of displacement measurement, but also greatly improves the calculation efficiency by focusing on the key feature points.

[0132] The present application also provides a displacement measurement device utilizing optical topological states, wherein the displacement measurement device utilizing optical topological states comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the displacement measurement method utilizing optical topological states in the above-mentioned embodiments.

[0133] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium, and when the instructions are executed on a computer, the computer executes the steps of the displacement measurement method using optical topological states.

[0134] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0135] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0136] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A displacement measurement method using an optical topological state, characterized in that: The displacement measurement method using optical topological states includes: The target object is illuminated by a preset laser emitting device, the optical signal is collected by an optical sensor, and the optical signal is converted into a digital signal; removing background noise from the digital signal to obtain a clean digital signal; Performing topological feature point identification on the clean digital signal to obtain a topological feature point set; specifically comprising: performing image reconstruction on the clean digital signal to obtain reconstructed image data; performing key part segmentation on the reconstructed image data to obtain multiple key part data; performing local contrast enhancement on the reconstructed image based on each key part data to obtain enhanced image data; performing feature point position identification on the enhanced image data to obtain a feature point position set; performing image segmentation on the enhanced image data based on the feature point position set to obtain multiple initial feature point images; performing image screening on the multiple initial feature point images to obtain multiple target feature point images; performing feature point descriptor construction on each target feature point image to obtain a descriptor set; based on the descriptor set, performing image pair matching on the multiple target feature point images through a preset matching algorithm to obtain multiple groups of image pairs; performing initial topological feature point screening on the multiple groups of image pairs to obtain an initial feature point set; performing deduplication processing on the initial feature point set to obtain the topological feature point set; Based on the topological feature point set, a displacement vector of the target object is calculated to obtain a target displacement vector.

2. The displacement measurement method using optical topological states according to claim 1, characterized in that: The method comprises: irradiating the target object with a preset laser emitting device, collecting light signals with an optical sensor, and converting the light signals into digital signals, including: The target object is irradiated by a preset laser emitting device, and at the same time, the object attributes of the target object are extracted to obtain the target object attributes; Based on the properties of the target object, collecting light reflection characteristic information of the target object; configuring the sampling frequency of the optical sensor according to the light reflection characteristic information, and collecting the light signal through the optical sensor; Calculating the average power of the optical signal to obtain the average power of the signal, and inputting the average power of the signal into a preset automatic gain control algorithm to perform input gain correction to obtain a target input gain; The target input gain is input into a preset analog signal converter for digital signal conversion to obtain the digital signal.

3. The displacement measurement method using optical topological states according to claim 2, characterized in that: The calculating the average power of the optical signal to obtain the average power of the signal, and inputting the average power of the signal into a preset automatic gain control algorithm to perform input gain correction to obtain a target input gain, comprises: The instantaneous power of the optical signal is calculated to obtain an instantaneous power set, wherein the expression of the instantaneous power is as follows: ;in, is the signal amplitude at time t, is the instantaneous power; The instantaneous power set is subjected to sliding average processing to obtain the signal average power, wherein the sliding average calculation formula is as follows: ; Where M is the time window length, is the average signal power, i is the time index value; Based on the signal average power, the signal power standard deviation of the optical signal is calculated to obtain the signal power standard deviation, wherein the signal power standard deviation calculation formula is as follows: ;in, is the signal power standard deviation; Based on the signal power standard deviation and the signal average power, a gain adjustment factor is calculated for the optical signal using a gain adjustment factor calculation formula to obtain a gain adjustment factor value, wherein the gain adjustment factor calculation formula is as follows: ;in, is the gain adjustment factor value, is the preset target power data, is the adjustment coefficient; A gain adjustment value is calculated based on the gain adjustment factor value to obtain a target gain adjustment value, wherein the gain adjustment value calculation formula is as follows: ;in, is the target gain adjustment value, is the conversion factor value; An input gain correction is performed based on the target gain adjustment value to obtain a target input gain.

4. The displacement measurement method using optical topological states according to claim 1, characterized in that: The step of removing background noise from the digital signal to obtain a clean digital signal comprises: Performing spectrum recognition on the digital signal to obtain a background noise frequency band; Based on the background noise frequency band matched filter, a plurality of initial filters are obtained; Calculating the cutoff frequency of each of the initial filters to obtain a cutoff frequency set; Based on the cutoff frequency set, parameter configuration is performed on each of the initial filters to obtain a plurality of target filters; Decomposing the digital signal to obtain multiple sub-signals; Performing signal recombination on the plurality of sub-signals to obtain a recombined signal; The background noise of the recombined signal is removed based on the multiple target filters to obtain the clean digital signal.

5. The displacement measurement method using optical topological states according to claim 1, characterized in that: The step of calculating the displacement vector of the target object based on the topological feature point set to obtain the target displacement vector includes: Performing cross-frame matching on the target object to obtain multiple consecutive frame images; Using the topological feature point set, image displacement point recognition is performed on multiple frames of the continuous frame images to obtain multiple groups of displacement points; Based on the multiple groups of displacement points, a displacement vector of the target object is calculated to obtain a target displacement vector.

6. A displacement measurement device using an optical topological state, characterized in that: The displacement measurement device using optical topological states comprises: The acquisition module is used to illuminate the target object through a preset laser emitting device, collect light signals through an optical sensor, and convert the light signals into digital signals; A removal module, used for removing background noise from the digital signal to obtain a clean digital signal; The identification module is used to identify the topological feature points of the clean digital signal to obtain a topological feature point set; specifically comprising: reconstructing the image of the clean digital signal to obtain reconstructed image data; segmenting the reconstructed image data into key parts to obtain multiple key part data; based on each key part data, locally enhancing the contrast of the reconstructed image to obtain enhanced image data; identifying the position of feature points on the enhanced image data to obtain a feature point position set; segmenting the enhanced image data based on the feature point position set to obtain multiple initial feature point images; performing image screening on the multiple initial feature point images to obtain multiple target feature point images; constructing feature point descriptors for each target feature point image to obtain a descriptor set; based on the descriptor set, performing image pair matching on the multiple target feature point images through a preset matching algorithm to obtain multiple groups of image pairs; performing initial screening of topological feature points on the multiple groups of image pairs to obtain an initial feature point set; performing deduplication processing on the initial feature point set to obtain the topological feature point set; The calculation module is used to calculate the displacement vector of the target object based on the topological feature point set to obtain the target displacement vector.

7. A displacement measurement device using an optical topological state, characterized in that: The displacement measurement device using optical topological states includes: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the displacement measurement device using optical topological states to perform the displacement measurement method using optical topological states according to any one of claims 1 to 5.

8. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the displacement measurement method using optical topological states as described in any one of claims 1 to 5 is implemented.

Images