Cable oil film uniformity detection system based on machine vision

Abstract

The invention relates to the field of cable detection, and discloses a cable oil film uniformity detection system based on machine vision, and the system comprises an active photoelectric illumination module which is used for generating a modulation optical signal of which the polarization state changes under a preset time sequence, and projecting the modulation optical signal to the surface of a cable; the synchronous photoelectric sensing and conversion module is used for receiving the modulated optical signal reflected by the surface of the cable and synchronously converting the modulated optical signal into a first electrical signal flow and a second electrical signal flow; wherein the first electrical signal flow represents multi-spectral intensity information of the reflected modulated optical signal, and the second electrical signal flow represents multi-polarization intensity information of the reflected modulated optical signal. By utilizing the sensitivity difference of the thin film interference model and the elliptical polarization model to the optical parameter change of the oil product, the systematic measurement error caused by batch replacement of the oil product or temperature fluctuation can be actively corrected, and the accuracy and long-term stability of the detection result are improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of cable detection, and particularly relates to a cable oil film uniformity detection system based on machine vision. BACKGROUND

[0002] In the field of modern industrial manufacturing, especially in the production process of metal cables, optical fibers, steel or precision parts, accurate thickness and uniformity online quality monitoring of the functional coating (such as lubricating oil film, antirust oil film, insulating paint layer, etc.) on the surface of the product is a key link to ensure the performance of the product and the smooth progress of the subsequent processing technology. Among many detection technologies, the non-contact measurement method based on optical principles has been widely applied due to its high speed, non-destructive and online integration advantages.

[0003] At present, there are many technical solutions for online detection of the coating state of the cable surface. Common methods are spectrum reflection method and elliptical polarization method. The spectrum reflection method calculates the thickness by emitting wide spectrum light to the surface of the cable and analyzing the interference characteristics in the reflected spectrum; the elliptical polarization method calculates the thickness by emitting a light beam with a specific polarization state and measuring the change of the polarization state after the light is reflected by the surface of the cable. Both methods can realize accurate measurement of film thickness in the range of nanometers to microns in principle.

[0004] However, when the above high-precision optical measurement method is independently applied to a dynamic industrial production line environment, a difficult-to-overcome technical bottleneck is encountered. Whether it is the spectrum reflection method or the elliptical polarization method, the thickness solving algorithm in the background depends on a pre-set, accurate physical model, one of the core parameters of the model is the optical constant (mainly the refractive index) of the measured oil film material. When the system is initially calibrated, a known refractive index value of a standard oil sample is used to establish the calculation model. However, in continuous industrial production, due to fluctuations in the ambient temperature or slight differences in the chemical components of the oil between different production batches, the refractive index of the oil film actually coated on the cable will drift, deviating from the initial calibration value.

[0005] When such refractive index drift occurs, the theoretical model based on fixed parameters of the system no longer matches the physical reality, which will directly lead to a systematic deviation in the thickness value calculated by the measurement algorithm. Since the system itself cannot perceive that the optical parameter of the oil has changed rather than the thickness has really changed, it cannot identify and automatically correct this deviation online. This potential measurement inaccuracy reduces the long-term reliability of the detection result, and when the measurement data is used for closed-loop process control of the production line, it may even produce false guidance to the production process. SUMMARY

[0006] In view of the deficiencies of the prior art, the application provides a cable oil film uniformity detection system based on machine vision, which solves the problem that the existing measurement method has high precision but is affected by material changes, thereby restricting the stability and accuracy.

[0007] To achieve the above object, the application is implemented by the following technical scheme: a cable oil film uniformity detection system based on machine vision, comprising: An active photoelectric illumination module is used to generate a modulated light signal with a change in polarization state under a preset timing and project it onto the surface of the cable. A synchronous photoelectric sensing and conversion module is used to receive the modulated light signal reflected by the surface of the cable and synchronously convert it into a first electrical signal stream and a second electrical signal stream. The first electrical signal stream represents the multispectral intensity information of the reflected modulated light signal, and the second electrical signal stream represents the multi-polarization state intensity information of the reflected modulated light signal. A data processing unit is in communication connection with the active photoelectric illumination module and the synchronous photoelectric sensing and conversion module. The data processing unit is internally configured with a logic circuit. Based on a physical model, the first electrical signal stream and the second electrical signal stream are processed to obtain information representing the thickness of the cable oil film and generate an output signal representing the uniformity of the cable oil film.

[0008] According to the above technical scheme: through the synchronous photoelectric sensing and conversion module, at the same time from the same point on the surface of the cable, two different physical dimension information are obtained in parallel: the multispectral intensity information representing the thin film interference effect and the multi-polarization state intensity information representing the change in polarization state of the reflected light.

[0009] Subsequently, the data processing unit receives the two strictly synchronized data streams and uses them to drive two independent physical models for thickness calculation. Since the sensitivity of the two models to the refractive index change of the oil film is different, when the oil property drifts, there will be a clear difference between the calculation results of the two channels. The application uses this difference signal as a real-time indicator of system error, and through a specific logic, selects a set of model parameters that best match the actual optical parameters of the current oil from a preset multiple model parameter library for calculation, thereby actively eliminating the systematic measurement error caused by the change of material parameters.

[0010] Preferably, the logic circuit in the data processing unit is further configured to: In the first processing circuit, the first electrical signal stream is processed according to a thin film interference model to obtain a first oil film thickness value; and in the second processing circuit, the second electrical signal stream is processed according to an elliptical polarization model to obtain a second oil film thickness value.

[0011] Preferably, the logic circuit further comprises a difference operation circuit configured to perform point-by-point subtraction between the first oil film thickness value and the second oil film thickness value to generate a difference signal representing systematic bias.

[0012] Preferably, the logic circuit is configured to call a new set of model parameters from the memory and adjust the physical model with the new set of model parameters when the amplitude of the difference signal triggers a preset condition. Preferably, the memory stores a plurality of sets of model parameters corresponding to different optical parameters of the oil.

[0013] Preferably, the logic circuit is configured to call a new set of model parameters that minimizes the amplitude of the difference signal.

[0014] Preferably, the logic circuit further comprises a signal fusion circuit configured to perform weighted average on the oil film thickness values calculated according to the adjusted physical model to generate a final oil film thickness value.

[0015] Preferably, the data processing unit is further configured to map the oil film thickness value into a pseudo-color image signal and calculate and output a digital signal representing statistical characteristics of the pseudo-color image signal, the digital signal including mean value and standard deviation.

[0016] Preferably, the data processing unit further comprises a Fourier transform processing module configured to process a one-dimensional sample data sequence extracted from the oil film thickness value to obtain a frequency spectrum signal thereof and identify a peak frequency component in the frequency spectrum signal as an output signal representing periodic non-uniformity.

[0017] Preferably, the synchronous photoelectric sensing and conversion module comprises: a light splitting device configured to split the reflected modulated light signal into a first light path and a second light path; a multi-wavelength filtering device configured in the first light path and configured to generate a first electrical signal stream; a fixed polarization detection device configured in the second light path and configured to generate a second electrical signal stream.

[0018] A machine vision-based cable oil film uniformity detection method, comprising the following steps: generating a modulated light signal with changing polarization state under a preset timing and projecting it onto the surface of the cable; receiving the reflected modulated light signal from the surface of the cable and synchronously converting it into a first electrical signal stream and a second electrical signal stream; wherein the first electrical signal stream represents the multi-spectral intensity information of the reflected modulated light signal, and the second electrical signal stream represents the multi-polarization state intensity information of the reflected modulated light signal; calculating a first oil film thickness value based on the first electrical signal stream and a second oil film thickness value based on the second electrical signal stream; calculating a difference signal between the first oil film thickness value and the second oil film thickness value, and adjusting a physical model used to calculate the first oil film thickness value and the second oil film thickness value; determining a final oil film thickness value based on the adjusted physical model, and generating an output signal representing cable oil film uniformity.

[0019] The present application provides a machine vision-based cable oil film uniformity detection system. The following advantages are provided: 1. The present application uses the difference between the thickness values calculated by the two independent channels as feedback to diagnose and select the physical model parameters pre-stored in the memory that are more matched with the current optical parameters of the oil product, so as to actively correct the systematic measurement error caused by oil product batch replacement or temperature fluctuation, thereby improving the accuracy and long-term stability of the detection results.

[0020] 2. The present application sets up a synchronous photoelectric sensing and conversion module to split and synchronously convert the collected reflected light signals into a first electrical signal stream representing multi-spectral intensity and a second electrical signal stream representing multi-polarization state intensity, ensuring that the information of two different physical dimensions is obtained from the same point on the cable surface at the same time, and providing accurate data basis for subsequent closed-loop calibration using data redundancy.

[0021] 3. The present application provides a multi-dimensional and quantitative evaluation result of oil film uniformity by generating an oil film thickness topological graph, calculating its statistical indicators, and analyzing its spatial frequency spectrum using a Fourier transform module. Through such design, the overall distribution, dispersion degree and existence of periodic non-uniformity defects caused by process equipment of the oil film thickness can be displayed. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 The figure is a schematic diagram of the system module architecture of the present application; Figure 2 The figure is a schematic diagram of the method steps of the present application. DETAILED DESCRIPTION

[0023] The technical solutions of the present application will be described clearly and completely below in combination with the drawings of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0024] Please refer to the drawings of the present application Figure 1 - the drawings of the present application Figure 2The embodiment of the present application provides a cable oil film uniformity detection system based on machine vision, which comprises: An active optoelectronic illumination module is used to generate a modulated light signal with changing polarization state under preset timing and project it onto the surface of the cable. A synchronous optoelectronic sensing and conversion module is used to receive the reflected modulated light signal on the surface of the cable and synchronously convert it into a first electrical signal stream and a second electrical signal stream. The synchronous optoelectronic sensing and conversion module comprises: a light splitting device for splitting the reflected modulated light signal into a first light path and a second light path; a multi-wavelength filtering device arranged in the first light path for generating the first electrical signal stream; and a fixed polarization detection device arranged in the second light path for generating the second electrical signal stream.

[0025] The first electrical signal stream represents the multi-spectral intensity information of the reflected modulated light signal, and the second electrical signal stream represents the multi-polarization state intensity information of the reflected modulated light signal. A data processing unit is in communication connection with the active optoelectronic illumination module and the synchronous optoelectronic sensing and conversion module. The data processing unit is internally configured with a logic circuit. The first electrical signal stream and the second electrical signal stream are processed based on a physical model to obtain information representing the thickness of the cable oil film and generate an output signal representing the uniformity of the cable oil film. In the first processing circuit, the first electrical signal stream is processed based on a thin film interference model to obtain a first oil film thickness value. In the second processing circuit, the second electrical signal stream is processed based on an elliptical polarization model to obtain a second oil film thickness value. The logic circuit further comprises a difference operation circuit for point-by-point subtraction of the first oil film thickness value and the second oil film thickness value to generate a difference signal representing systematic deviation. When the amplitude of the difference signal triggers a preset condition, the logic circuit calls a new set of model parameters from the memory and adjusts the physical model with the new set of model parameters. The memory stores a plurality of sets of model parameters corresponding to different optical parameters of oil.

[0026] The logic circuit specifically calls a new set of model parameters that can minimize the amplitude of the difference signal. The logic circuit is also provided with a signal fusion circuit for weighted average of the oil film thickness values calculated based on the adjusted physical model to generate a final oil film thickness value. The data processing unit is also used to map the oil film thickness value into a pseudo-color image signal and calculate and output a digital signal representing the statistical characteristics of the pseudo-color image signal, including the mean value and the standard deviation. The data processing unit further comprises a Fourier transform processing module for processing a one-dimensional sampling data sequence extracted from the oil film thickness value to obtain a frequency spectrum signal and identifying a peak frequency component in the frequency spectrum signal as an output signal representing periodic non-uniformity.

[0027] Specifically, the function of the active optoelectronic illumination module is to generate a modulated light signal with periodically changing polarization state at a preset timing, and project it onto the moving cable surface. Inside the module, there is a broadband light source, a collimating and expanding system, a rotating linear polarizer driven by a servo motor, and a cylindrical mirror. The light beam generated by the broadband light source is collimated and expanded, and then passes through the rotating linear polarizer. The broadband light source can be a halogen lamp or a xenon lamp to provide a continuous spectrum covering the visible to near-infrared band. In some cases where higher spectral power density is required for specific materials, a supercontinuum laser can also be used as a broadband light source. The data processing unit sends control signals to the servo motor to drive the rotating linear polarizer to rotate at a preset angular velocity or angular sequence (e.g., P1, P2, …, P n ) to periodically change the polarization direction of the transmitted light.

[0028] Finally, the cylindrical mirror reshapes the circular spot with modulated polarization state into a linear spot, which is projected and covers the cross-section of the cable to be measured. The function of the synchronous optoelectronic sensing and conversion module is to receive the modulated light signal reflected from the cable surface and synchronously convert it into a first electrical signal stream and a second electrical signal stream.

[0029] The core of this module is a light splitting device, such as a 50:50 non-polarizing beam splitting cube. This light splitting device splits the modulated light signal carrying the oil film information reflected from the cable surface into a first light path and a second light path with equal light intensity.

[0030] In the first light path, there is an imaging lens group, a multi-wavelength filtering device, and a first image sensor (such as a CMOS or CCD camera). The multi-wavelength filtering device can be a rotating filter wheel loaded with multiple narrowband filters with different center wavelengths (e.g., λ1, λ2, …, λ m ) Under the control of the data processing unit, the filter wheel is synchronized with the exposure of the first image sensor, and a frame of image is collected at each wavelength in turn. This series of images constitutes the first electrical signal stream, whose data content represents the multi-spectral intensity information of the reflected light.

[0031] In another embodiment, the multi-wavelength filtering device can also use a mechanically motionless electrically controlled tunable filter, such as a liquid crystal tunable filter (LCTF) or an acousto-optic tunable filter (AOTF). When using such filters, the data processing unit can switch wavelengths by applying different driving voltages or radio frequency signals, achieving faster wavelength switching speed.

[0032] In the second light path, another imaging lens group, a fixed polarizing device and a second image sensor are configured, the fixed polarizing device is a linear polarizer with fixed optical axis direction, and the exposure time of the second image sensor is strictly synchronized with the angular position of the rotating linear polarizer in the active photoelectric illumination module. When the polarization angle of the illumination light is P1, P2, …, Pn respectively, a group of images are synchronously acquired by the second image sensor. n

[0033] The series of images constitute a second electrical signal stream, and the data content thereof represents the multi-polarization state intensity information of the reflected light. The data processing unit is the control and calculation core of the system, and can be implemented by an industrial computer, a digital signal processor or a field programmable gate array.

[0034] The first and second electrical signal streams are received from the synchronous photoelectric sensing and conversion module through a standard industrial camera interface, and in an embodiment, the standard industrial camera interface adopts CameraLink or GigEVision. Synchronous control instructions are sent to the active photoelectric illumination module and the multi-wavelength filtering device through a control interface. The data processing unit is internally configured with a logic circuit for executing the method of the application.

[0035] In combination with the system structure described above, the detection method of the application is described in detail.

[0036] When the system is running, the active photoelectric illumination module emits linear light with a preset polarization angle sequence P k . After being reflected on the cable surface, the light signal enters the synchronous photoelectric sensing and conversion module. In the module, the first image sensor acquires a multi-spectral intensity image Iλ j (x, y) under different filters λ j , and at the same time, the second image sensor acquires a multi-polarization intensity image I k (x, y) corresponding to each illumination polarization angle P Pk . The entire acquisition process is accurately controlled by the synchronization signal of the data processing unit, ensuring the correspondence of all images in space and time.

[0037] After the data processing unit receives the two groups of image data, the internal logic circuit starts parallel processing.

[0038] In the first processing circuit, the first electrical signal stream, i.e., the sequence of multi-spectral intensity images, is processed. First, the original light intensity image Iλ j is converted into an absolute reflectance image Rλ j through a pre-calibrated system response function.

[0039] ​Then, for each pixel (x, y) in the image, a measurement vector [Rλ1, Rλ2,..., RλN] is formed, consisting of the multispectral reflectance values of the pixel. m The circuit retrieves from internal memory a pre-generated spectral look-up table (LUT-Spec) based on thin-film interference model. The look-up table stores the correspondence between different oil film thickness d and theoretical reflectance vectors. By searching the look-up table for the entry that best matches the measurement vector, the first oil film thickness value d spec (x, y) for the pixel is obtained.

[0040] In the second processing circuit, the second electrical signal stream (sequence of multi-polarization intensity images) is processed. For each pixel (x, y) in the image, the light intensity values I k at multiple polarization angles P Pk are used to solve for the normalized ellipsometric coefficients (a, b) that characterize the polarization state change by Fourier analysis or least-square fitting. The circuit retrieves from internal memory a pre-generated polarization look-up table (LUT-Pol) based on ellipsometric model. The look-up table stores the correspondence between different oil film thickness d and theoretical ellipsometric coefficients. By searching the look-up table for the entry that best matches the measurement coefficients (a, b), the second oil film thickness value d pol (x, y) for the pixel is obtained.

[0041] Then, adaptive physical model calibration is performed. A difference operation circuit receives the first oil film thickness map d spec (x, y) and the second oil film thickness map d pol (x, y), and performs pixel-wise subtraction to generate a difference signal map e(x, y) = d spec (x, y) - d pol (x, y). Then, the statistical mean value m ε of the difference signal map in a pre-defined region of interest (ROI) is calculated. The absolute value |m ε | of the mean value is compared with a pre-defined threshold.

[0042] The threshold is zero or a value close to zero, which represents the normal measurement fluctuation of the system. The threshold can be determined by repeatedly measuring a standard sample at the initial calibration of the system, and analyzing the statistical distribution of the difference signal map e(x, y). For example, the threshold can be set as three times the standard deviation of the distribution. If |m ε | is greater than the threshold, the model adaptive calibration procedure is triggered.

[0043] Once the calibration procedure is triggered, the data processing unit accesses its internal memory. This memory stores a physical model library consisting of multiple lookup table pairs {(LUT-Spec,i,LUT-Pol,i)}, where each lookup table pair is based on a specific assumed value n of the oil film refractive index. 1,i The calibration logic, calculated sequentially or via a preferred gradient descent method, selects different lookup table pairs to repeat the thickness calculation process in step two, and calculates the mean value |μ| of the new difference signal obtained in each calculation. ε,i |

[0044] Ultimately, the system will automatically determine and select the value that makes the mean of the difference signal |μ ε,i The minimized lookup table pair (LUT-5pec,opt,LUT-Pol,opt) serves as the currently calibrated physical model for subsequent calculations. This process updates the oil film optical parameters used by the system in real time.

[0045] After the physical model completes adaptive calibration, the system uses a defined lookup table to recalculate the calibrated thickness map. and

[0046] A signal fusion circuit performs weighted fusion of the two calibrated thickness maps to generate the final oil film thickness map d. final (x,y). The fusion formula is: Weighting coefficient w spec and w pol It is pre-calibrated based on the signal-to-noise ratio characteristics of the two methods in different thickness ranges, and satisfies w spec +w pol =1.

[0047] Specifically, the weighting coefficient w spec and w pol It is not a fixed value, but rather a function w that can be configured as the thickness. spec (d) and w pol (d). For example, in regions with thinner oil film thickness, the signal-to-noise ratio of elliptic polarization is higher, and w can be set... pol >w spec In regions with thicker oil films, the measurement results obtained by thin-film interferometry are more stable, and w can be set. spec >w pol .

[0048] Finally, the data processing unit uses the final oil film thickness map d final (x,y) generates a quantitative, uniform output signal. In the spatial domain, the average thickness μ of the entire image is calculated. d and thickness standard deviation σ d, as an indicator of the overall thickness and the degree of dispersion of the oil film.

[0049] In the frequency domain, a Fourier transform processing module extracts the spatial frequency spectrum of the first electrical signal stream from a series of consecutive acquisitions of the first electrical signal stream. final The one-dimensional thickness profile data is extracted from the image. A fast Fourier transform (FFT) is performed on the one-dimensional data to obtain its spatial frequency spectrum. By performing peak detection on the frequency spectrum, the non-uniformities at certain spatial frequencies caused by factors such as periodic vibrations of the equipment can be identified, and the peak frequency and amplitude can be output as part of the output signal.

[0050] All the above steps are automatically performed within the data processing unit, thus realizing online, adaptive, and quantitative cable oil film uniformity detection.

[0051] In another embodiment of the present application, the synchronized optoelectronic sensing and conversion module can adopt a single-camera scheme with higher integration. Specifically, a “snapshot multispectral polarimetric image sensor” can be used. Such a sensor integrates a miniature optical array on top of its pixel array, which is periodically arranged with filter elements of different center wavelengths and polarimetric elements of different polarization directions.

[0052] In this embodiment, the synchronized optoelectronic sensing and conversion module can simultaneously obtain multispectral intensity information and multi-polarization state intensity information from different functional pixels of the image sensor within a single exposure.

[0053] After the data processing unit receives the raw image containing mixed information, it first performs a demosaicing algorithm to re-register and separate the data from different functional pixels, thus reconstructing the first electrical signal stream, i.e., the multispectral image sequence, and the second electrical signal stream, i.e., the multi-polarization image sequence, which are functionally equivalent to the aforementioned dual-camera embodiment.

[0054] The subsequent processing steps, including dual-channel independent solving, adaptive physical model calibration, thickness fusion, and uniformity evaluation, are the same as in the aforementioned embodiment. This single-camera scheme simplifies the optical path structure and volume of the system.

[0055] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements, and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A machine vision-based cable oil film uniformity detection system, characterized by, The application relates to a cable oil film thickness measurement system, comprising: an active optoelectronic illumination module for generating a modulated light signal with changing polarization state at preset timing and projecting the modulated light signal onto the surface of a cable; a synchronous optoelectronic sensing and conversion module for receiving the modulated light signal reflected by the surface of the cable and synchronously converting the modulated light signal into a first electrical signal stream and a second electrical signal stream; wherein the first electrical signal stream represents the multispectral intensity information of the reflected modulated light signal, and the second electrical signal stream represents the multiple polarization state intensity information of the reflected modulated light signal; a data processing unit in communication connection with the active optoelectronic illumination module and the synchronous optoelectronic sensing and conversion module, wherein the data processing unit is internally provided with a logic circuit, the first electrical signal stream and the second electrical signal stream are processed based on a physical model to obtain information representing the thickness of the cable oil film, and an output signal representing the uniformity of the cable oil film is generated.

2. The machine vision-based cable oil film uniformity detection system of claim 1, wherein: The logic circuit in the data processing unit is further configured to: in a first processing circuit, the first electrical signal stream is processed based on a thin film interference model to obtain a first oil film thickness value; in a second processing circuit, the second electrical signal stream is processed based on an elliptical polarization model to obtain a second oil film thickness value.

3. The machine vision-based cable oil film uniformity detection system of claim 2, wherein: The logic circuit further comprises a difference operation circuit for point-by-point subtraction of the first oil film thickness value and the second oil film thickness value to generate a difference signal representing systematic deviation.

4. The machine vision-based cable oil film uniformity detection system of claim 3, wherein: When the amplitude of the difference signal triggers a preset condition, the logic circuit calls a new set of model parameters from a memory and adjusts the physical model with the new set of model parameters; wherein the memory stores a plurality of sets of model parameters corresponding to different optical parameters of oil products.

5. The machine vision-based cable oil film uniformity detection system of claim 4, wherein: The logic circuit specifically calls a new set of model parameters that can minimize the amplitude of the difference signal.

6. The machine vision-based cable oil film uniformity detection system of claim 3, wherein: The logic circuit is further provided with a signal fusion circuit for weighted average of the oil film thickness values calculated based on the adjusted physical model to generate a final oil film thickness value.

7. The machine vision-based cable oil film uniformity detection system of claim 1, wherein: The data processing unit is further used for mapping the oil film thickness value into a pseudo-color image signal and calculating and outputting a digital signal representing the statistical characteristics of the pseudo-color image signal, wherein the digital signal comprises an average value and a standard deviation.

8. The machine vision-based cable oil film uniformity detection system of claim 6, wherein: The data processing unit further comprises a Fourier transform processing module for processing a one-dimensional sampling data sequence extracted from the oil film thickness value to obtain a frequency spectrum signal and identifying a peak frequency component in the frequency spectrum signal as an output signal representing periodic non-uniformity.

9. The machine vision-based cable oil film uniformity detection system of claim 1, wherein: The synchronous optoelectronic sensing and conversion module comprises: a light splitting device for splitting the reflected modulated light signal into a first light path and a second light path; a multi-wavelength filtering device arranged in the first light path for generating the first electrical signal stream; a fixed polarization detection device arranged in the second light path for generating the second electrical signal stream.

10. A machine vision-based cable oil film uniformity detection method, according to the machine vision-based cable oil film uniformity detection system of any one of claims 1-9, characterized in that, The application further relates to a cable oil film thickness measurement method, comprising the following steps: generating a modulated light signal with changing polarization state at preset timing and projecting the modulated light signal onto the surface of a cable; receiving the modulated light signal reflected by the surface of the cable and synchronously converting the modulated light signal into a first electrical signal stream and a second electrical signal stream; wherein the first electrical signal stream represents the multispectral intensity information of the reflected modulated light signal, and the second electrical signal stream represents the multiple polarization state intensity information of the reflected modulated light signal; calculating a first oil film thickness value based on the first electrical signal flow and a second oil film thickness value based on the second electrical signal flow; calculating a difference signal between the first oil film thickness value and the second oil film thickness value, and adjusting a physical model used to calculate the first oil film thickness value and the second oil film thickness value; determining a final oil film thickness value based on the adjusted physical model, and generating an output signal characterizing uniformity of the cable oil film.

Images